Manuel d'utilisation / d'entretien du produit bh5700 du fabricant ZNYX
Aller à la page of 359
HP bh5700 ATCA 14-Slot Blade Ser ver Ethernet Switch Blade First Edition Manufacturing Part Number: AD171-9603A June 2006.
Ethernet Switch Blade User's Guide release 3.2.2j page ii.
Legal Notices The information in this document is sub ject to change without notice. Hewlett-Packard makes no warranty of any kind with regard to this manual, including, but not limited to, the implied warranties of merchanta bility and fitness for a particular purpose.
About the Ethernet Sw itch Blade Manual This manual includes everything you need to begin usi ng the HP Ethernet Switch Blade with OpenArchitect software, Release 3.
Table of Contents Chapter 1 Overview of the Ethernet Switch Blade ........................................................... 17 High Performance Embedded Switching...................................................................... 17 Advanced TCA® Compliant.
Rapid Spanning Tree................................................................................................ 50 To Enable Rapid Spanning Tree:......................................................................... 51 Port Path Cost...........
Chapter 5 Fabric Switch Administration........................................................................... 73 Setting the Root Password............................................................................................ 73 Adding Additional Users.
Example Configuration Scripts............................................................................ 92 Overview of OpenArchitect VLAN Interfaces.................................................... 93 Tagging and Untagging VLANs..................
Classical Targets................................................................................................ 111 ZNYX Targets................................................................................................... 112 ZACTION Examples.
SNMP and OpenArchitect Interface Definitions............................................... 134 ifStackTable Entries........................................................................................... 135 SNMP Configuration.....................
Booting the Duplicate Flash Image ............................................................................ 159 Chapter 13 Network Configuration Problems ............................................................... 160 Interface Overview........
Chapter 17 Restoring the Factory Default Configurat ion................................................ 188 Chapter 18 Before Calling Support..................................................................................189 Appendix A Fabric Switch Command Man Pages.
zgr................................................................................................................................ 297 zgvrpd.............................................................................................................
Figure 6.3: Init Script Flow................................................................................................ 86 Figure 7.1: Multiple VLANs..............................................................................................94 Figure 7.
Ethernet Switch Blade User's Guide release 3.2.2j page 15.
Ethernet Switch Blade User's Guide release 3.2.2j page 16.
Chapter 1 Overview of the Ethernet Swi tch Blade The Ethernet Switch Blade is a 72-port Adva ncedTCA® Hub and providing Gigabit Ethernet. Up to 14 ATCA node boards may be addressed via the PICMG 3.
OpenArchitect Switch Management The OpenArchitect software component – open sourc e Linux, IP protocol stack, control applications and the OA Engine – runs on two embedde d PowerPC microprocessors. OpenArchitect provides extensive manage d IP routing protocols and other open standards for switch management.
Ethernet Switch Blade P ort Configuration Base switch Quick Reference ShelfManager1 zre22 ShelfManager2 zre13 ISL channel ( Base node2 ) zre23 Base nodes 3-14 zre0-11 Base nodes 15,16 zre 20-21 Front .
You will find the Ethernet Switch Blade ha s a straightforward installation and configur ation. UNIX or Linux system management skills a nd some understanding of network protocols will be required. Configure the Ethernet Switch Blade s to your networking application before you begin using the OpenArchitect switch.
network-enabled Linux implementation. The purpose of the routing table is to tell the packet forwarding software where to forward the data packets. In Linux, the packet-forwa rding algorithm is operated in software.
Figure 1.2: OpenArchitect Softwar e Structure OpenArchitect applications are used to progr am and configure the Ethernet Switch Blade. These applications are implemente d above the libraries and RMAPI. Ethernet Switch Blade User's Guide release 3.
Chapter 2 Port Cabling and LE D Indicators The PICMG 3.1 standard defines an embedded Ethe rnet environment for Telco chassis. This environment includes two switch fabric sl ots that create a dual star Ethernet net work to the fourteen node slots.
4. Reinsert the switch into the shelf chassis and power up. Use a terminal emulation program to access the switch console. Out of Band Ports (OOB P orts) Each switch, fabric and Base, in a Ethe rnet Switch Blade unit has out-of-band (OOB) Etherne t ports on the front panel.
Ethernet Switch Blade User's Guide release 3.2.2j page 25 Figure 2.1: LED Reference.
Ethernet Switch Blade User's Guide release 3.2.2j page 26.
Chapter 3 High Availability Networkin g High availability networking is achieve d by eliminating any single point of failure through redundant connectivity: Redundant cable s, switches and network interfaces for hardware, combined with HA software solutions on both the hos ts and switches to control the HA hardware and maintain connectivity.
VRRP Since most end nodes use default router addres ses, the change of the default router address during a switch failover would require the end nodes to reconfigure. Layer 3 switches that failover must maintain the default router address to ma intain the end node's IP transparent failover.
Switch Replacement and Reconf iguration When a switch fails, it must be repl aced. The replacement switch will likely requi re proper configuration. For transparent switch replac ement, the newly replaced switch must le arn its configuration from its Surviving Partner.
The configuration and runtime scripts creat ed are as follows: • S70Surviving_partner Switch initialization script that is run at boot time. This script will restart the switc h with the original configuration given to zspconfig . Optionally, zspconfig will run this script from the i nitial invocation.
When using a Linux Bonding driver on the node card, the bonding dri ver should be configured for Mode 1 (active/standby). See the Linux Bonding docum entation at http://sourceforge.
sibling_addresses: zhp1 = 10.0.0.30, 10.0.0.31 netmask 255.0.0.0; Now configure the virtual address for each s ibling group. We are going to create a virtual interface across one VLAN, but not for the inte rconnect. This provides a single point to connect/route to the VLANs.
#vrrp_mode: block_crossconnect; The next sections determines t he failover mode between the Surviving Partner switches. Ther e are three modes: • switch - Failover by switch. Failover from Maste r switch to Backup on any port failure. The switch with the most links becomes the new Master.
#start_script:/etc/rcZ.d/SxxScript; #start_script:/etc/rcZ.d/SyyScript; # vrrpd_script: Allows the user to add scripts to be executed during # vrrpd state transitions. These scripts are run from the end of the # /etc/rcZ.d/surviving_partner/vrrpd.script file.
Once the configuration files are comple te, run the zspconfig utility on the Master to configure all the scripts: NOTE: This command can take 60 seconds or more with n o screen output. zspconfig –f zsp.conf You will see output similar to this: zspconfig -f zsp.
# This script will likely need modification for your particular # network setup. # # In this example the Egress ports, zre20..23 and zre48..50 are # not managed by HA since how, or if, these ports are managed by HA is # dependent on the external devices they are connected to.
zconfig zre0, zre4, zre8, zre12, zre16, zre24, zre28, zre30, zre32, zre34, zre36, zre38, zre40, zre42 = untag1; zconfig zre1, zre5, zre9, zre13, zre17, zre25, zre29, zre31, zre33, zre35, zre37, zre39,.
vrrp_virtual_address: zhp1 = 10.0.0.42 netmask 255.0.0.0; vrrp_virtual_address: zhp2 = 11.0.0.42 netmask 255.0.0.0; vrrp_virtual_address: zhp3 = 12.0.0.42 netmask 255.0.0.0; vrrp_virtual_address: zhp4 = 13.0.0.42 netmask 255.0.0.0; # Port definitions # Define to what the ports are connected.
# crossconnect ports of the VRRP Backup. The block_crossconnect mode is # meant as a replacement for STP, however, the switches connected to the # crossconnect ports must be Ethernet Switch switches running Surviving Partner.
failover_mode: port; # VRRP_msg_rate is the time in milliseconds between transmissions # VRRP messages on the interconnect. The VRRP protocol requires the # absence of 3 VRRP messages before concluding that the remote switch # has failed. The msg_rate must match the msg_rate of all siblings.
# Fabric portions of the 7100 switch. The actual coordination is dependent on the # setting of the board_synchronization_mode and the failover_mode. In # switch failover_mode the number of up links in both switch planes is # considered. In vlan and port failover mode they are not.
# gated_template: Allows the user to provide a template for the # gated.conf file to be used by the sibling group. #gated_template: /etc/rcZ.d/surviving_partner/gated.
Finally, it lets the currently saved S70Surviving_Partner script execute. This case would be the case of a power up of an already configured backup sw itch when the other HA switch is unavailable. This case could occur after losing power to the entire chassis.
"zsp.primary.conf"; } host SECONDARY { fixed-address 100.0.0.31; option dhcp-client-identifier "SECONDARY"; option vendor-encapsulated-options "zsp.secondary.conf"; } } The zsp.primary.conf and zsp.secondary.conf files must be placed in the tftp location on the machine, often /tftpboot .
request vendor-encapsulated-options; require vendor-encapsulated-options; The last step is to modify the start up scripts that run zspconfig to use the -c option. The -c option allows you to provide a dhclient.conf script rather then having zspconfig create a default.
Chapter 4 Fabric Switch Configu ration Two switches, two consoles There are two separate switch port ions in the Ethernet Switch Blade units, the base s witch and the fabric switch. The fabric switch handles the data traffic for the ATCA rack over ports 0-47.
Changing the Shell Prompt You may use standard bash shell procedures to change the prompts on your base switches. Many sites choose a system that distingui shes among the individual switches at their locati on. The same rules apply for saving your choice ( zsync) a s for all other configuration changes.
Overview of OpenArchitect VLAN In terfaces A zhp device is associated with one VLAN. zhp ma y have one or more physical ports and their associated zre devices. A VLAN from the viewpoint of the switch is a logical mapping of ports based on intended use.
Switch Port Interfaces For each switch port, OpenArchitect c reates a separate interface with its own MAC a ddress called a ZNYX raw Ethernet ( zre ). After the initial power up, 48 zre interfaces are created, one for each in band port. You cannot directly acce ss or modify the zre interfaces.
ifconfig zhp1 0.0.0.0 # # At this point the system will act as a Layer 2 switch # across all ports. Also, the system will accept telnet() # connections on 10.0.0.43 on any port. Script(s) may then # be run to reinitialize the system and modify its # configuration.
To Enable Rapid Spanning Tree: Create a VLAN containing the ports that will be a part of the Linux bridge running Rapid Spanning Tree. This example will use ports 0-3 (unta gged): zconfig zhp0: vlan1=zre0.
Layer 3 Switch Configuration The previous section outlines the Layer 2 switch configuration that is automatically confi gured when you initially bring up the OpenArchitec t switch. In order to communicate between Layer2 interfaces, you must properly setup routing.
In the S50layer3 script separate VLANs are set up for ea ch port. The VLANs, are labeled as zhp0..zhpn . Each VLAN is associated with an individua l zre interface.
the number of IP addresses as applicable. In the exam ple below, the IP address is changed for the interface in the ifconfig command line of the scr ipt. From: ifconfig zhp0 10.0.0.43 netmask 255.255.255.0 broadcast 10.0.0.255 up To: ifconfig zhp0 193.
interface 10.0.1.42 passive interface 10.0.2.42 passive . . . interface 10.0.13.42 passive interface 10.0.14.42 passive interface 10.0.15.42 passive • Defines the netmask used in the interface . define 10.0.0.43 netmask 255.255.255.0; define 10.0.1.
. . interface 10.0.13.43 ripin ripout version 1; interface 10.0.14.43 ripin ripout version 1; interface 10.0.15.43 ripin ripout version 1; • Imports routes learned through the RIP protocol. import proto rip { all; }; • Exports all directly connected routes and routes l earned from the RIP protocol.
Or for OSPF: cp /etc/rcZ.d/examples/S55gatedOspf /etc/rcZ.d cp /etc/rcZ.d/examples/gated.conf.ospf /etc/rcZ.d • Open and make configuration changes to the listed conf file to coincide with the current Layer 3 configuration (that is, adjust IP addresses and numbe r of interfaces available).
Marking and Re-marking The OpenArchitect switch can mark or rem ark packets using the TOS field or 802.1p tag. This is also controlled through the Linux iptables utili ty. Scheduling The servicing of configured queues by the switc hing fabric is referred to as scheduling.
you may want to move your set of iptables commands t o a start up script to run upon initialization. This could be a ccomplished by creating a standalone " S " script and plac ing that script into / etc/rcZ.d . Restrictions on Implementation Several restrictions exist on the r ules that can be implemented on the FFP hardware.
On the other hand, in the following sequence of rule s, the position of the rule that drops SYN packets is important. Since the set of fi elds it examines is not a subset of the fields exa mined by the ACCEPT rules, and visa versa, the order ing rule given above does not apply.
By default, INPUT, FORWARD and OUTPUT chains are insta lled on boot up. Additional rules can be installed for the other chains . Additionally, one can write software extensions to add mor e chains. Figure 4.2 provides an illustration of the Firewall Flow.
send to CPU action is specified, it is sent to the INPUT chain for further processing. If there is no valid way to forward the packet, it is dropped. If t he switch is configured to forward the packet, it is sent to the FORWARD chain. Next the hardware FORWARD chain is walked.
The type can be preceded by ! to match a ny message except the type listed, for example, -- icmp-type ! 1 Specifying TCP or UDP po rts If the protocol is TCP or UDP, the -s ( or --sport ) and -d (or --dport ) opti ons specify the TCP or UDP ports to match.
--drop Drops the packet --accept Accepts the packet --set-prio <val> Set the 802.1p priority to <val> --use-prio <val> Use queue priority <val> --copy-cpu Send the packet to the CPU.
FORWARDING Chain supports all of them. tc and zqosd tc , which stands for Traffic Control, is a mec hanism for enabling Quality of Service on Linux. tc uses three functional objects: queuing dis cipli.
qdisc pfifo 100: dev zhp0 limit 32p The tc command is applied to a device, so dev zhp0 must be specified. Note that a VLAN, such as zhp0 , and a port, such as zre0 , are eac h treated as devices. Breakdown of the options: handle 100:0 Defines the handle for the queuing discipl ine.
The byte-limited FIFO queue case differs only s lightly from the packet-limited FIFO case. T he syntax is almost identical. In hardware the limit is based on 128-byte cells. The speci fied byte limit is divided by 128 to determine the cel l limit. Always specify a byte limit of at least 128 bytes to avoid setting the queue length to zer o.
index of the list element (numbering f rom 0) and q is the value specified by that element. So, thi s example would read: Priority 0 maps to Queue 1 Priority 1 maps to Queue 2 Priority 2 maps to Queue 2 Priority 3 maps to Queue 2 Priority 4 maps to Queue 3 Note that the tc priority map applies to a 4- bit field.
The U32 Filter The U32 filter provides the capability to mat ch on fields in the L2, L3 or L4 header of a packet. Each match rule gives the locati on of the field to be tested, which is always a 32 bit word, a mas k selecting the bits to be test ed, and a value which is to be matched by the packet field.
Although the translation rules handle some inconsistency between software and hardware, a use r must define a combination of rules that is reasonable in hardware, to ensure predictable re sults. Handle Semantics All examples have illustrate d zqosd copying tc rules into hardware.
• The PDP sends that policy to the PEP. • The PEP installs the policy and applie s it to future traffic. As long as COPS is running, a connection between the P EP and PDP should stay open. A PEP could query a PDP at any time asking for a policy deci sion.
The pepd utility requires a PDP that has impl emented the above RFCs and drafts. Until all dra ft standards are approved, the certain COPS-PR data types will not be assigned OIDs.
Chapter 5 Fabric Switch Administration One of the main benefits of the OpenArchite ct switch is that it runs Linux, so much of the switch administration is already familiar to most network or system administrators.
Enter new password: Re-enter new password: Password changed. ZX7100-OA<release no.># zsync ZX7100-OA<release no.># Setting up a Default Route If you wish to access the switch from some pla ce other than a directly attached network, you may want to setup a default route.
dhcpd Consult Linux Network administration m anuals for more information on DHCP and configuration options. To use DHCP to set your IP addresses automatical ly on boot up, uncomment the the following line in /etc/init.
/sbin/rpc.statd /usr/sbin/rpc.mountd -r Once the above servers are started, you c an mount a remote NFS file system. mount rhost:nfs_file_system local_mount_point If the remote NFS file system you’re mounting is on an OA switch, you should mount with caching disabled.
Now start nfsd to export the mount points and begin answer ing requests from remote clients. /sbin/rpc.nfsd –r To export file systems automatic ally on boot, edit /etc/init.d/rcS , uncomment the /sbin/rpc.nfsd command line by removing the #. /sbin/rpc.
SNMP Agent Simple Network Management Protocol (SNMP) is the defacto standard for network management. An SNMP agent maintains a structure of data for a network device in a virtual information database, called a Management Inf ormation Base (MIB).
Supported MIBs RFC 2573: S NMP Applications RFC 2574: U ser-based Security Model (USM) for version 3 of the Simple Network Management Protocol (SNMPv3) RFC 2575: V iew-based Security Model (VACM) for .
Supported Traps SNMPv2-MIB: coldStart SNMPv2-MIB: authenticationFailure IF-MIB: linkUp IF-MIB: linkDown UCD-SNMP-MIB: ucdShutdown RMON-MIB: risingAlarm RMON-MIB: fallingAlarm VRRP: vrrpTrapNewMaster VRRP: vrrpTrapAuthFailure EGP (rfc1213): egpNeighborLoss BGP4-MIB: bgpEstablished BGP4-MIB: bgpBackwardTransition Table 5.
Link and SNMP Status Physical Link Status SNMP Operational Status zre1 zre2 zre1 zre2 zhp0 down down down down down down u p down up up up down up down up up up up up up Table 5.3: Link and SNMP Status The administrative status i s directly controlled by ifconfig up/down.
response. The processing for link up and link down traps is now user c onfigurable. As the default, traps conform to RFC2863, meaning the trap conte nts will include: ifIndex, ifAdminStatus and ifOper.
mirrored (copied and transmitted) to por t 12. This mirroring would be in addition to any Layer 3 or Layer 2 switching. zmirror zre0 zre12 zmirror zre1 zre12 zmirror zre2 zre12 To clear the current mirroring use the -t option. The -e option can be used to indicate that packets being sent on a given port should be copi ed to the mirror_to port.
Chapter 6 Fabric Switch Maintenance This chapter includes basic informa tion about the OpenArchitect switch environment incl uding an overview of the file system structure, modi fying and updating switch files, upgrading the switch driver and kernel, and implementing a s ystem recovery.
Figure 6.2: Boot Flow Chart Under normal circumstances, the booting up proc ess follows the process outlined in Figure 6.2. During boot up, the zmon bootloader reads the device boot string to locate and validate the correct application image to l oad.
Saving Changes Any modifications made to the scripts for your particular configuration must be properly saved or your changes are lost when you reboot. The file syste m for the switch only exists in memory. A rewritable overlay is contained within t he upper four megabytes of the first application f lash.
configuration files contained in / e t c / r c Z . d In order to telnet into the box, there must be a configured interface with a proper IP address. For e xample, zhp0 is configured with the IP address 10.
zsync /etc/hosts • Reboot the system. System Hangs During Boot After attaching the system console cabl e, if the system hangs during boot, try booting with the –i option as described in the previous section. It is possible that important Linux system fil es became corrupted and incorrectly save d in the flash overlay.
Download the OpenArchitect image to a local system. The OpenArchitect image is very cl ose to the limit of free space available on a default system so you may need to clear some space prior to downloadi ng the OpenArchitect image to the switch. Check for free space with the df c ommand.
Using apt-get apt-get is a utility created by the Debia n Linux community to allow remote fetching and installation of software stored in a repository in Debian package format. It allows users to keep their software up-to-date with the latest binaries, and install new software without the need to recompile.
Chapter 7 Base Switch Configuration At this point, the OpenArchitect Ether net Switch Blade should be installed and powered up for the first time. This chapter helps you conne ct and configure the base switch by presenting command line examples as well as a disc ussion of the example configuration scripts .
files into flash for reloading. Changing the Shell Prompt You may use standard bash shell procedures to change the prompts on your base switches. Many sites choose a system that distingui shes among the individual switches at their locati on. The same rules apply for saving your choice ( zsync ) as for al l other configuration changes.
• S50multivlan - Script which sets up multiple untagged VLANs. The first VLAN includes the first ten 10/100/1000 ports, the next contains the last ten 10/100/1000 ports, the third VLAN contains two 10/100/1000 ports, the last VLAN contains the last two 10/100/1000 ports.
Tagging and Untagging VLANs The OpenArchitect switch is capabl e of switching VLAN tagged and untagged data packets. VLAN tagged packets conform to the 802.1q specific ation and the packet header contains an additional four bytes of VLAN tag information.
Figure 7.2: Layer 2 Switch During the initial power up, a startup script called /etc/rcZ.d/S50layer2 is executed at boot time creating a single untagged VLA N (IP interface labeled as zhp0 ) which includes all Ethernet and gigabit ports as one Layer2 switch.
Using the S50layer2 Script The S50layer2 script can be used and example, or edited to customiz e your Layer2 setup. For example, to reconfigure the IP address on your Layer 2 switch , • Open the S50Layer2 file in the Linux vi editor. • Change the IP address value listed under the Linux ifconfig(1M) command line.
brctl show brctl showbr bzhp0 Port Path Cost Each port has an associated cost that c ontributes to the total cost of the path to the Root Bridge when the port is the root port. The smaller the cost, the better the path. The Ethernet Switch Blade uses the following IEEE 802.
zconfig zhp1: vlan2=zre5..8 zconfig zre5..8=untag2 Now, use ifconfig to assign each zhp interface an IP address, ifconfig zhp0 10.0.0.1 ifconfig zhp1 11.0.0.1 At this point, the Linux host has enough informati on to route between the networks of the directly attached interfaces, 10.
Linux IP zre1 VLAN 2 zre4 zre3 VLAN 3 zre2 VLAN 4 zre11 VLAN 11 zre7 VLAN 7 zre9 VLAN 9 zre6 VLAN 6 zre8 VLAN 8 zre10 VLAN 10 zre5 VLAN 5 VLAN 14 zre12 VLAN 12 zre13 VLAN 13 zhp0 - zhp2 3 zre20 VLAN 1.
• Runs the OpenArchitect zl3d . The zl3d application monitors the Linux routing tables and updates the switch routing tables for each inte rface configured above. /usr/sbin/zl3d zhp0..23 zl3d initially creates and adds eac h zhp interface (VLAN) to the switch routing tables.
• VLAN 4, zhp3: for last set of six ports, zre18- zre23 Each VLAN interface is labeled z h p N in the file, where N is a value from 0-3. Each interface is untagged and assigned its own IP address (see Figure 7.
(10.0.0.42-10.0.3.42), assigns the netmask and brings them up. ifconfig zhp0 10.0.0.42 netmask 255.255.255.0 broadcast 10.0.0.255 up ifconfig zhp1 10.0.1.42 netmask 255.255.255.0 broadcast 10.0.1.255 up ifconfig zhp2 10.0.2.42 netmask 255.255.255.0 broadcast 10.
example): • Starts GateD with Rip1 using gated.conf.rip1 as the confi guration file: /usr/sbin/gated –f /etc/rcZ.d/gated.conf.rip1 The GateD conf file specifies the f ollowing configuration comman.
interface all noripin noripout • Opens sending and receiving packets for selected int erfaces. interface 10.0.0.42 ripin ripout version 1; interface 10.0.1.42 ripin ripout version 1; interface 10.0.2.42 ripin ripout version 1; . . . interface 10.0.13.
cp /etc/rcZ.d/examples/gated.conf.rip1 /etc/rcZ.d Or for RIP2: cp /etc/rcZ.d/examples/S55gatedRip2 /etc/rcZ.d cp /etc/rcZ.d/examples/gated.conf.rip2 /etc/rcZ.d Or for OSPF: cp /etc/rcZ.d/examples/S55gatedOspf /etc/rcZ.d cp /etc/rcZ.d/examples/gated.conf.
Marking and Re-marking The OpenArchitect switch can mark or rem ark packets using the TOS field or 802.1p tag. This is also controlled through the Linux iptables utili ty. Scheduling The servicing of configured queues by the switc hing fabric is referred to as scheduling.
Running zfilterd Before starting zfilterd , ztmd must be running. Your can star t both from within a script, or directly from the command line . For example, ztmd zfilterd iptables rules can be entered at any time.
action that will take place . For example, the rules: iptables -a FORWARD -i zhp3 -j DROP iptables -a FORWARD -i zhp3 -o zhp1 -p tcp --dport smtp -j ACCEPT result in SMTP packets received on any port in zhp3 to be sent for any port in zhp1 ; all other packets from zhp3 would be dropped.
Introduction Firewall rules are stored in tables. The se tables are sometimes also known as firewal l chains or just chains . Tables normally store rules for what are known as hooks , which can be looked as packet-path junctions. There are five de fined hooks: PRE-ROUTE, POST-ROUTE, INPUT, OUTPUT and FORWARDING.
Packet Walk When a packet comes in via one of the int erface ports, the base switch makes a routi ng decision. If the packet was destined for the base switch itself or if the send to CPU action is speci fied, it is sent to the INPUT chain for further processing.
--icmp-type ping The type can be preceded by ! to match a ny message except the type listed, for example: --icmp-type ! 1 Specifying TCP or UDP po rts If the protocol is TCP or UDP, the -s ( or --sport ) and -d (or --dport ) opti ons specify the TCP or UDP ports to match.
ZNYX Targets ZACTION This is the ZNYX Action target. Parameters for ZACTION: --drop Drops the packet --accept Accepts the packet --set-prio <val> Set the 802.1p priority to <val> --use-prio <val> Use queue priority <val> --copy-cpu Send the packet to the CPU.
Extensions to the default matches These are described in the Linux packet filtering HOWTO at: http://netfilter.org/docume ntation/index.html#documentation-howto ZNYX FORWARDING Chain supports all of them. tc : Traffic Control The switch supports up to eight queues for each por t, including the cpu port.
queue number + 1 after the qdisc handle. So the highest priority queue in this example is 105:8. NOTE: 16 values must be provided for the priomap list. T his is a feature of the Linux priority system, which uses 16 priority level s. The last eight values given will be ignored.
handle 100:0 Defines the handle for the queuing discipl ine. This handle may be used to reference the pfifo queue. Note that the ha ndle is included with the output of the qdisc ls command. (100:0 and 100: are equi valent in tc .) The choice of handle is significant for zqosd .
match ip tos 0xa0 0xe0 would match an IP precedence of 5. Specific fields can also be speci fied by giving their offset from the beginning of the IP hea der and a field name of u8 , u16 , or u32 , depending on the width of the fiel d.
tc qdisc add dev zre1 ingress //ingress qdisc for zre1 tc qdisc add dev zhp2 ingress //ingress qdisc for vlan The filter add command changes slightly, the pa rent is now a special handle ffff:fff1, so using the same filter as the first example : tc filter add dev zre1 parent ffff:fff1 protocol ip u32 match ip dst 10.
omitted, and the packet is not dropped, the e gress queue will be determined by the priority of t he packet, either from the 802.1p priority for ta gged packets or the default priority for untagged packets for the ingress port.
specified numerically for eithe r out-of-profile or in-profile actions. The numeric val ue is a decimal integer action code shown in the table below. If the action requires a paramet er, the parameter value is multiplie d by 256 and added to the action code.
for a u16 match. In many cases, there is a fiel d name that can be used for the match, elimina ting the need to specify the offset. U match selectors Field Match Equivalent ip src a.
OpenArchitect switch though, because the norm al case is for packets to be switched in hardware. For that reason, zqosd must be used to shadow tc configurati on into hardware. Like zfilterd , zqosd works with ztmd , which provides the actual hardwar e interaction.
In tc , the prio queuing discipline establishes multi ple queues and specifies their associate d priority map. Although WRR support is not part of the standard tc distribution, it has been added to the prio disciplinE. The following example illustrat es WRR.
many packets sent as queue 0, queue 2 will have four times as many, and queue 3 will have six times as many. wrr parameters a re scaled such that the maximum value is no more t han 15.
tc filter add dev zhp0 protocol arp parent 100:0 u32 match u32 2 0xffff at +4 flowid 100:30 Combining Queuing Discip lines Any of the queue length limiting discipline s can be used with the bandwidth management queue disciplines, by defining them with the handle of one of the classes as their parent.
Figure 7.6: COPS Network Architecture A PDP contains all of the policy rulers for it s associated PEPs. A PDP typically stores rules in a data and is a dedicated server, not a f orwarding device. A PEP is any network device that has to enforce policy decisions.
and relaying those requests to its PDP. By contrast, the provisioning model is based on l onger lasting policy. The expectation is that policy should be administratively defi ned at the PDP and pushed to the PEPs as needed. OpenArchitect is a COPS-PR client.
where, PDP address: The IP address of the PDP. Default is loopback (127.0.0.1) PDP port: The destination port on which to open a COPS connection. Default is 3288. PEPID: The PEP Identifier Role-If: A mapping of roles to interfaces. The name of the role is followed by a comma-delineated list of interfaces.
Chapter 8 Base Switch Administration One of the main benefits of the OpenArchite ct switch is that it runs Linux, so much of the switch administration is already familiar to most network or system administrators.
ZX6000-OA<release no.># zsync ZX6000-OA<release no.># Setting up a Default Route If you wish to access the switch from some pla ce other than a directly attached network, you may want to setup a default route. Use the rout e command to set a default gateway.
Network Time Protocol ( NTP) Client Configuration NTP is a protocol for setting the real time clock on a system. There are numerous primary and secondary servers available on the network. F or more NTP information, and a list of available NTP servers, see the following URL: http://www.
All the necessary servers are included in /etc/init.d/rcS but are commented out by default. To automatically start all NFS client services each time you boot, uncomment t he NFS Client servers. Go to the /etc/init.d/rcS file. Uncomment the following command lines by removing the # sign.
ftpd Server Configuration The switch itself can also be configured to run a n FTP server ( ftpd ). See the Linux Reference Guide for details of the ftpd command. You will ne ed to add a user to the switch in order to connect via ftp from a remote host, since root is not allowed ftp access.
Supported MIBS RFC 1155: Structure and Identification of Management Information for TC P/IP-based internets RFC 1227: SNMP MUX Protocol and MIB RFC 1493: Definitions of Managed Objects for Bridges (ob.
Supported MIBS ZNYX Networks Private MIB Custom ZNYX MIB to support software and hardware features not covered by standard MIBs. The Private MIBs are ZX7100BASE.MIB AND ZX7100FABRIC.MIB, pointed to by ZNYX-H.MIB. UCD-SNMP Enterprise MIB UCD-SNMP MIB related to management and monitoring of the LINUX ho st Table 8.
status is down, then the operational status will be down independent of the underlying link state. You must ifconfig up the zres to see the operational link status for a zre . When the administrative status is up, the oper ational status is dependent on the underlying physical state.
IMPORTANT: For NET-SNMP agents, these objects ( sysLocation.0, sysContact.0 and sysName.0 ) ordinarily are read-write. However , specifying the value for one of these objects by giving the appropriat e token in snmpd.
zmirror mirror_from mirror_to After executing the following three c ommands, packets received on ports 0, 1 and 2 would be mirrored (copied and transmitted) to por t 12. This mirroring would be in addition to any Layer 3 or Layer 2 switching. zmirror zre0 zre12 zmirror zre1 zre12 zmirror zre2 zre12 To clear the current mirroring use the -t option.
Chapter 9 Base Switch Maintenance This chapter includes basic informa tion about the OpenArchitect switch environment incl uding an overview of the file system structure, modi fying and updating switch files, upgrading the switch driver and kernel, and implementing a s ystem recovery.
Under normal circumstances, the booting up proc ess follows the process outlined in Figure 6-2. During boot up, the zmon bootloader reads the device boot string to locate and validate the correct application image to l oad.
Saving Changes Any modifications made to the scripts for your particular configuration must be properly saved or your changes are lost when you reboot. The file syste m for the switch only exists in memory. A rewritable overlay is contained within t he upper four megabytes of the first application f lash.
Booting with the –i option If you cannot telnet into the switch and Linux fails to boot, it is likely that a change saved by zsync has left the switch in an inaccessible state.
System Hangs During Boot After attaching the system console cabl e, if the system hangs during boot, try booting with the –i option as described in the previous section. It is possible that important Linux system fil es became corrupted and incorrectly save d in the flash overlay.
the limit of free space available on a default system, so you may need to clear some space prior to downloading the new OpenArchitect im age to the switch.
Using apt-get apt-get is a utility created by the Debia n Linux community to allow remote fetching and installation of software stored in a repository in Debian package format. It allows users to keep their software up-to-date with the latest binaries, and install new software without the need to recompile.
Chapter 10 Connecting to the Ethernet Switch Bl ade The Ethernet Switch Blade has two compl etely separate switching subsystems within one ATCA blade supporting both Base Interface and F abric Interfa.
console port. An RS-232 to RJ-45 adapter is required. Fabric Interface Hub System : A 48-port Gigabit Ethernet Switch that pr ovides PICMG 3.1 Option 2 (2.0 Gb/s) Ethernet service for a full 14-slot ATCA chassis. All connect ors for the fabric interface hub and it’s processor ar e labeled “fabric”.
To attach the console cable to the Ethernet Swit ch Blade switch: 1. Plug the RJ-45 end of the console cable (P/N 6900-63006, shippe d with the HP bh5700 ATCA 14-Slot Blade Server) into the RJ-45 Consol e Port (1) on the front panel. 2. Connect the DB-9 end of console cable into a s tandard Modem Eliminator Cable (normally locally available).
NOTE: The OOB port is not active by default with the factory c onfigured configuration. The first time you log into the switch either in-band or thr ough the console cable you must use the ifconfig command to make the port active.
9. Reinsert the switch into the system and power up. 10. Use a terminal emulation program to access the switch console. Fabric Interface Out of Band Eth ernet Connection Connect an Ethernet cable from the Ethernet S witch Blade front panel MGMT OOB (4 in Figure 10.
Chapter 11 Diagnosing a Failed E thernet Switch Blade Activation The Ethernet Switch Blade must tra nsition through a series of states (M0–M4) to become act ive in an ATCA shelf. After the Ethernet Switc h Blade has reached the M4 state, it will becom e active and start the boot process of the Ope nArchitect Switch Management environment.
FRU State HotSwap LED Status Healthy LED Status Solution M0 OFF OFF No power. Board not inserted correctly. 1. Remove and re-insert board. 2. If board does not power-up after re-insertion, try a different slot.
FRU State HotSwap LED Status Healthy LED Status Solution switch through a console cable. If OpenArchitect is running, and abnormal be havior is occurring, please see Network Configurati on Problems for information on network issues.
sensor information. Examine the System E vent Log (SEL) on the ShMM and determine if critica l sensor events have been logged for the switch in quest ion. If the switch has reported critical sensor data for temperature or voltage, the ShMM ca n prevent it from booting.
clia board -v 7 or clia board -v 8 These commands generate an output that reports if the ShMM thinks it has granted access to ports on the switches. Check the Shelf Manager User’s Guide for the expected output. Ethernet Switch Blade User's Guide release 3.
Chapter 12 Troubleshooting a Failed OpenArchitect Load The OpenArchitect operating system is loaded fr om the FlashROM memory into RAM when the Ethernet Switch Blade is activated by the Shelf Manager.
The Ethernet Switch Blade is equipped with a Random Acce ss Memory (RAM) disk and three Read-Only Memory (ROM) devices, including, a boot ROM and two appli cation flash Ethernet Switch Blade User's Guide release 3.
The boot ROM is located on device 0 and contai ns the OpenArchitect zmon application that operates as a boot loader and include s a device bootstring. Device 1 contains the application fl ash 1 image of the Linux operating system and t he OpenArchitect overlay file system.
properly attach the console cable. Booting Without the Overlay File If you cannot telnet into the switch and Linux fails to boot, it is likely that a change saved by zsync has left the switch in an inaccessible state.
If the switch still is unable to boot, s ee Booting the Duplicate Flash Image , below. Booting the Duplicate Flash Image Another recovery method, if Linux fails to boot, is to temporarily boot the factory-installed duplicate image located in the second fl ash device.
Chapter 13 Network Configuration P roblems Many reported problems on a booted switch wil l ultimately be traced back to user errors in the layer 2 or layer 3 switch configuration. In some c ases, symptoms from an improperly configured switch can masquerade as potential ha rdware problems.
Physical Slot 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Fabric Port 3 19 11 - - 3 7 15 27 Fabric 51** * Base Interface Inter-Switch Link (I SL) ** 10 Gigabit Ethernet Fabric Interfac e - Update Channel Table 13.
2. S30e1000 - Script that loads the e1000 driver module for the Out-of- Band Ethernet ports. (Editing this script is not recommended.) S40vpd - Script that checks the current OA version, and loads into t he Vital Product Data (VPD) area if necessary. (Editing this script is n ot recommended.
Default Fabric Interface Configuration Editing the S50layer2 script can change the Ethernet Switch Blade Fabric Interface default configuration. The S50Layer2 script and included example scr ipts ( /etc/rcZ.d/examples ) can be used as templates to create custom sc ripts.
ifconfig Default Screen Output for the Bas e Interface Ethernet Switch Blade User's Guide release 3.2.2j page 164 Figure 13.2: Linux Networking Environ ment Interfaces [ZX7100-OA3.2.2h]# ifconfig lo Link encap:Local Loopback inet addr:127.0.0.1 Mask:255.
Configuration Troubleshooti ng Problem Solution No Connection Physical Link problem. Check to see if the port LED is lit. If the LED port is not lit, then you may have a bad cable connection. OR Configuration Error. Connect through the console port (See Chapter 10).
The following table will translate the zlc output t o link status. Link Port Status Link Speed Pause Faults OK Zre (x) EKEY_DISA BLED Auto Enable Internal Fault EKEY_ENAB LED 1000fd UP 1000hd DOWN 100.
10hd – Ethernet Half Duplex Pause: Enable: a port that can temporarily suspend the data transmission between two network devices in the event t hat one of the devices becomes congested. Pause enabled devices can reduce bottlenecks by making the network more efficient.
NOTE: this is the zlc output for a single Etherne t Switch Blade Base Interface in the default configuration with no line cards installed in the chassis. Querying Fabric Interface ekey Status Link Status for a single port To query a link status for a single port type zre<x> query.
Ethernet Switch Blade User's Guide release 3.2.2j page 169 [ZX7100-OA3.2.2h]# zlc zre0..51 query zre0: <EKEY_DISABLED, AUTO, PAUSE ENABLE, EXT_FLT ON, OK ON> zre1: <EKEY_DISABLED, AUTO, .
Network Connectivity Troubleshooting No Connection If the port LED is lit on the front panel, the switch has established a physical connection and the problem is a network configuration error. Check t o see if both devices are configured to be on the same network (ex.
have an active remote device att ached, then first bring down the ports which do not have active connections expected to make sure there is a legitimate EXT FLT condition. If loss of communications is suspected on an e xternally wired port, make sure to check and test affected cables.
Traceroute Test It’s possible to trace a network path usi ng the traceroute command. The following is an example of a Layer 2 traceroute with only t wo devices. Ethernet Switch Blade User's Guide release 3.2.2j page 172 sh-2.04# traceroute 192.
Chapter 14 Isolating Hardware Failures 1. Fl ash 10. Switch Chip (U69) 2. EEPROM 11. Zone 3 ATCA Connector 3. PH Y 12. Isolation Transformers 4. C PU 13. 4-port PHY 5. SDRAM 14. Zone 2 ATCA Connector 6. I solation Transformer 15. Zone 1 ATCA Connector 7.
1. Isolation Transformer 2. Zone 3 ATCA Connector 3. Isolation Transformer 4. Switch Chip (U60) 5. SDRAM 6. Switch Chip (U59) 7. Isolation Transformer Ethernet Switch Blade User's Guide release 3.
1. Isolation Transformer 8. Flash ROMs 2. 4 Port PHY 9. FPGA 3. CPU (U22) 10. ZMC Connector 4. 10 Gigabit XFP 11. Zone 3 ATCA Connector 5. 10 Gigabit PHY 12.
6. Isolation Transformer 13. Isolation Transformers 7. Power Supply 14. 4 Port PHY Hardware Subsystem In the following tables, refer to the ident ified component-area numbers on indicated in the pictures in the proceeding section. The indi cations of malfunction may be identified eit her during normal operation, or in response to a specific tes t.
Base ZM C 0 # ZMC 1 # Hardware Subsystem Indications of Malfunction any of the following indications: • Error message via OpenArchitect due to inability to access the registers within the switch chip, or a failure of DMA transfers. • Loss of switch functionality, such as the inability to forward packets, or forwarding packets in error.
Duplicate Flash Image . If the switch can succ essfully boot from FlashROM device 2, then FlashROM device 2 is fully operational. Testing the Switch Fabric You can test the functionality of the swit ch fabric by running the zlc command. The zlc command outputs the link status for any Ethe rnet Switch Blade interface.
Example Output: NOTE: This is the zlc output for a single Ethern et Switch Blade Base Interface in the default configuration with no line cards installed in the chassis. Testing the onboard RAM You can test the onboard memory by running the free c ommand.
If the “Used” and “Free” memory statis tics do not add up to the Total memory, the software environment may have a memory leak caus ed by a software error. Reboot the switch. If the problem persists after a reboot. R un the top command to list the memory utiliz ation of all current processes.
To test the operational status of t he control processors you can do the following: Hardware Fault Connect to the console port of either the Bas e or Fabric Interface control processor (See Chapter 10 for more information). If you cannot communicate with the Ether net Switch Blade, the control processor may have encountered a software error.
INT FLT LED is illuminated, replace the switch and return it for repair. Ethernet Switch Blade User's Guide release 3.2.2j page 182.
Chapter 15 High Availability Trou bleshooting The ATCA environment will usually contain a high-a vailability failover configuration betwee n two ATCA switches in the chassis.
Chapter 16 Switch Firmware Overview There are three components to the fir mware on the Ethernet Switch Blade: 1. Bootloader firmware ( zmon ) 2. OpenArchitect firmware 3. IPMI firmware Some hardware and software problems can be res olved by updating the firmware to the latest version.
Key: PN: Base Interface Switch Assembly Number SN: Base Interface Switch Serial Numbe r V6: OpenArchitect Version Number VP: IPMI Firmware Version VZ: BootLoader Version Number The following output is shown for the 3.
Updating the Switch Firmware Currently, the OpenArchitect and bootloader components are the only upgradeable firmware on the Ethernet Switch Blade. Upgrading t he IPMI software is not currently supported. BootLoader Firmware Upgrade: 1. Download the bootloader image to a local system.
Surviving Partner daemons to think there is a failure , resulting in link oscillation. Base Interface: zflash -d 1 rdr6000.zImage.initrd Fabric Interface: zflash –d 1 rdr7100.zImage.initrd IPMC Firmware Upgrade: Upgrading the IPMC Firmware through OpenArchitec t is not currently supported.
Chapter 17 Restoring the Factory Default Configuration You should use this procedure if the contents i n Flash Device 1 are corrupt and you need to restore the switch to the factory defaul t configuration.
Chapter 18 Before Calling Suppo rt Because of the highly customized confi gurations that can be applied by customers to their ATCA switch environment, the focus must be on dat a collection to get a snapshot of the current switc h configuration and network traffic activi ty.
Figure 18.1: ROM Devices in OpenAr chitect The boot ROM is located on device 0 and contai ns the OpenArchitect zmon application that operates as a boot loader and includes a device bootstring. Device1 contains the applicati on flash1 image of the Linux operating system and the OpenArchitect overlay file system.
Appendix A Fabric Swit ch Command Man Pages OpenArchitect applications are im plemented above the OpenArchitect libraries and the R MAPI interface. OpenArchitect applicati ons are used for normal operation of the switch, for runtime status and diagnostics, and for prototyping new a pplications development.
vrrpconfig NAME vrrpconfig – Configure and control the running vrrpd SYNOPSIS vrrpconfig [-d <level>] -- <vrrpd parameters> vrrpconfig [-d <level>] [-k] [-a] [-p] [-s <vid>] DESCRIPTION vrrpconfig provides communication with a running vrrpd daemon.
EXAMPLES Here is an example of using the -- invocat ion method that changes the priority to 99 for the Virtual Router associated with the Virtual Router Identifier 1: vrrpconfig -- -v 1 –p 99 SEE ALSO vrrpd Ethernet Switch Blade User's Guide release 3.
vrrpd NAME vrrpd – Virtual Router Redundancy Protocol Daemon SYNOPSIS vrrpd -i ifname -v vrid [-f piddir] [-s] [-a auth] [-p prio] [-nhb] [-I ifname] [-d delay] [-m address] [-M ] [-B] [-S script] [-c conf_file] [-D level] ipaddr DESCRIPTION vrrpd is an implementation of Virtual Re dundant Routing Protocol (VRRPv2) as specified in RFC2338.
the –i option. -s Toggle preemption mode (Enabled by default). Preemption means that a Master switch will go to Backup if a current Backup has higher priority. -M Become MASTER when priority is equal. Be sure it is only set on one host or the switches will oscillate.
The result of which to use the native MAC address of the interface. Using the –n option is not recommended. -D <level> Set debugging output to the supplied level <ipaddr> the ip address(es) of the virtual server SEE ALSO vrrpconfig Ethernet Switch Blade User's Guide release 3.
zbootcfg NAME zbootcfg − Modifies the boot parameters of t he OpenArchitect switch. SYNOPSIS zbootcfg -a | -d <device number> [<boot_string>] DESCRIPTION zbootcfg is used to display or modify the boot parame ters on the switch.
application flash. Typically this is required before updating application flash 1. B y booting the alternative image, if a fail ure occurs during the programming of application flash 1, rec overy is easier. zbootcfg -d 2 The next example passes the -i option t o the booting kernel.
zconfig NAME zconfig - Configures the OpenArchitect switch. SYNOPSIS zconfig [-h <host_name>] [-d <level>] [-a] [-t] [{-f <file>} | <configuration>] DESCRIPTION zconfig creates Virtual Local Area Network (VLAN) groups of switch port s or trunks.
input. If the -f flag is not used, a single line of configuration data can be entered as parameters to zconfig . CONFIGURATION SYNTAX zconfig takes configuration data from standar d input or from a file with the -f option. In either case, the configuration syntax is the same.
ports. teardown Removes the trunk interface, making the ports which were part of the trunk available for configuration in other trunks or VLANs. all mac [ source_address | destination_address ] ip [ s.
The syntax for a network interface state ment is the interface name followed by a colon and a n action. Network interface state ments are used to create or tear down a VLAN group and can consist of one or a list of network interfac e names; followed by a colon and then an action.
named zhp1 . This VLAN includes four switc h ports, zre1 , zre10 , zre11 , zre13 . zhp0: vlan100 = zre1,zre10,zre11,zre13; The next statement adds two switch port s, zre1 , zre2 and zre3 , to an existing network interface and VLAN. zhp0: vlan100 = zre1.
zre1=untag1; If port 0 is also a member of VLAN 100, packets f or VLAN 100 are sent from this port with a VLAN tag as part of their header. In the next example, the switch ports 10, 11, and tr unk 2 are configured as untagged members of VLAN 100.
The following statement creates 14 VLAN groups with VID numbers 1-14. Each VLAN contains the same switch port, port 1, represente d as zre1 . zhp0 ..13: vlan1..14 = zre1; The plus (+) wildcard can be used with the las t port listed to auto-increment that port number before each VLAN group is created.
The previous configuration can be used for creat ing a 14 port Layer 2 switch, all 14 ports assigned to the same VLAN. SEE ALSO zl3d Ethernet Switch Blade User's Guide release 3.
zcos NAME zcos - class of service queue control SYNOPSIS zcos [-h <hostname>] [-d <level>] [ -u <default priority> ] [ -m q0,q1,q2,q3,q4,q5,q6,q7 ] [-n <queue length list in packe.
Each cos queue is limited in the number of packets it can hold waiting scheduling; the mem ory used by each queue is managed to provide a guarant eed space with additional space shared among all queues for a port. OPTIONS Most options are optionally followed by a <port list> , which may inc lude zre port ranges, like zre0.
Packets are first counted against the re served space for a queue. When that space is occupied, additional memory is used from the dynamic me mory pool until the dynamic pool usage limit f or the port is reached. Any additional packets received for the queue on this port are dropped.
-q all | qmap | qinfo | scheduler [<port list>] Queries the current COS/QOS Settings. all - Displays all of the queue mappings, queue limi ts, metering and scheduling settings qmap - Displays the priority to COS queue mappings . qinfo - Displays queue limits for the COS queues .
zdog NAME zdog - Configure and send heartbeats to watch dog enabl ed drivers. SYNOPSIS zdog [-d <level>] -h | -i <interval> | -n <heartbeats> zdog [-d <level>] -b zdog [-d <.
component of zdog insures that the CPU is functioni ng well enough to execute something. The software component of zdog when launched from monit insures that monit is running to perform higher level tasks. And finally monit can be used to monitor any or all critical system resources and processes in the system.
zfilterd NAME zfilterd - A daemon to use the filter hardware of the OpenArchitect switch for filtering based on iptables(8) rules. SYNOPSIS zfilterd [-d <level>] [-p <port>] [-f] [-l] [-i .
zflash NAME zflash − Loads images into the flash ROMs on the OpenArc hitect switch. SYNOPSIS zflash -d <dev> [-o|-O <offset>] <image_file> <upgradeipmi.img> DESCRIPTION zflash enables you to program the flash ROMs on the switch.
Exercise caution when using this command, a s an error can render your switch inoperable. Do not interrupt this process until complete. SEE ALSO zbootcfg Ethernet Switch Blade User's Guide release 3.
zl2, zl2mc, zl3host, zl3net, zvlan NAME zl2, zl2mc, zl3host, zl3mc, zl3net, zvlan – Formatted display of OpenArchitect generic tables. zl2 displays the abstraction API’s layer 2 ta ble. zl2mc displays the abstraction API’s layer 2 multi cast table.
display on the user console. The format of the output is table-dependent. Port mapping affects the ports referenced in the generic table s. (Ports listed in order from 1 to 15) Headers describing the column being displayed ar e printed after every 22 lines of output, which makes it easy to pipe through more(1).
The following command deletes the above ent ry: zl2 -c -m 00:c0:95:45:00:00 The following command displays all ent ries of the zl2 table: zl2 Be careful, the -c option does not ask. The fol lowing command deletes all entries in the zl2 table: zl2 -c SEE ALSO zal Ethernet Switch Blade User's Guide release 3.
zgvrpd NAME zgvrpd - GARP VLAN Registration Protocol (GVRP) daemon fo r the OpenArchitect switch. SYNOPSIS zgvrpd [-d <level>] [-f] [-h <hostname>] [-p <ppa>] [-t <target>] DESCRIPTION zgvrpd is run after the network interfaces are created and initialized with zconfig , and started with ifconfig(1M) .
OPTIONS -d <level> Sets the level of debugging output required by zgvrpd. The default level is zero (0). Setting the debug level higher produces more output. Five (5) is currently the maximum output level. -f Run zgvrpd in foreground. Default is to run it in background.
zl2d NAME zl2d - Layer 2 daemon for the OpenArchitect switch. SYNOPSIS zl2d [start | stop] [-t <msecs>] [-d <level>] [-f] [-p <priority>] <iface..> DESCRIPTION zl2d is run after the network interfaces are created and initialized with zconfig .
OPERATIONS zl2d manages the Spanning Tree state fi elds in the switch of each port within the bridge(s). Based on a timer, zl2d reads the port inform ation for each Linux bridge and updates the switch when necessary.
zl3d NAME zl3d - Layer 3 daemon for the OpenArchitect switch. SYNOPSIS zl3d [-h <host_name>] [-t <msecs>] [-b] [-e] [-l] [-n] [-d <level>] <iface ..> DESCRIPTION zl3d is run after the network interfaces are created and initialized with zconfig , and started with ifconfig(1M).
produces more output. Four (4) is currently the maximum output level. <iface…> The network interfaces on which zl3d should operate. These network interfaces must first be created by zconfig. zl3d does not operate with standard network interface cards.
zlc NAME zlc − link and LED control SYNOPSIS zlc [-h <hostname>][-d <level>][-x] <port_list> <action> [on | off ] zlc [-h <hostname>][-d <level>][-x] <action&g.
EXAMPLES In the following example, zlc forces the line speed of port 1 t o 100 Full duplex. The interface must be down to change the speed. Assuming zre1 i s part of interface zhp0 , ifconfig zhp0 down zlc zre1 100fd The external fault, internal fault , and ok LEDs can be set on a per port basis or globally .
SEE ALSO ifconfig(8) Ethernet Switch Blade User's Guide release 3.2.2j page 227.
zlmd NAME zlmd − monitor link changes or hot swap events. SYNOPSIS zlmd [-h <hostname>] [-b] [-d <level>] {-f <file>} | <configuration> DESCRIPTION The zlmd application is intended to run as a daem on, waiting for a configured event to occur and then running the program configured for that event.
<port-list> A list of ports in the same forms supported by zconfig, e.g. zre1,zre2 or zre10..14 <program> Path to an executable program or script to be run when the event occurs.
zlogrotate NAME zlogrotate − Rotates log files. SYNOPSIS zlogrotate [-b] [-t time] [-s segment size] [-n # of files] [-f file to rotate] DESCRIPTION zlogrotate rotates the selected file every [t ime] seconds if the file is larger than [se gment size].
zmirror NAME zmirror - Set packet mirroring on an ingress or egress port. SYNOPSIS zmirror -a | -t zmirror [-e] <from_list> <to_port> DESCRIPTION zmirror sets packet mirroring from a given set of ports to a given port . Turning on packet mirroring causes a copy of the packet to be se nt to the to port.
zmirror is cumulative: zmirror zre1, zre2, zre3 cpu Is the same as: zmirror zre1 cpu zmirror zre2 cpu zmirror zre3 cpu Setting a different to port will ove rwrite the previous setting and direct previously mi rrored ports to a new to port. Given the last setup the fol lowing will change port 1 traffic to be forwarded to port 10.
zmnt NAME zmnt − Expands the read/write files onto the RAM dis k. SYNOPSIS zmnt [-c] <directory> zmnt [-c] -t <file> zmnt [-c] –l DESCRIPTION zmnt expands files from flash onto the RAM disk that have been previ ously saved with zsync .
In the following example, zmnt the current overla y into a tar file called overlay.tar zmnt –t overlay.tar The resulting tar file can now be saved on a dif ferent host as a snapshot of the overlay at that point in time. Use zsync to restore the ove rlay on the switch: zsync –t overlay.
zpeer NAME zpeer – Application for High Availability comm unication between the Fabric and Data switches. SYNOPSIS zpeer [-d <level>] local|peer <command> <value>|query zpeer [-d &.
displayed by query, but must be set at initial ization. After setting the healthy state, the query will retur n the backup state. backup Used to reflect the backup state of vrrpd master Used to reflect the master st ate of vrrpd The priority value is a value betwee n 0 and 255.
be also reset. SEE ALSO zspconfig Ethernet Switch Blade User's Guide release 3.2.2j page 237.
zqosd NAME zqosd – monitors tc(8) commands to implement classification filters and queuing disci plines in hardware. SYNOPSIS zqosd [-d <level>] [-p <port>] [-f] [-l] [-i <pid>] [-.
SEE ALSO ztmd, tc(8), zfilterd Ethernet Switch Blade User's Guide release 3.2.2j page 239.
zrc NAME zrc - Packet rate control SYNOPSIS zrc -b | -m | -d | -t | -a [-p <port>] [-v <vlan>] [-g <group>] [-M <mac_addr>] [-T <timeout>] [-D <level>] <rate> DESCRIPTION zrc sets rate control on Broadcast, Multicast and/or Destination Lookup Failure (DLF) packets.
zreg NAME zreg - Read and write registers and tables on the OpenArchite ct switch switching hardware. SYNOPSIS zreg [-p <ppa>] [-w] [-i <index>] [-t <index>] [-k] [-h <hostname>.
default zreg configures the OpenArchitect switch that is locally connected (i.e., the one that is on the local PCI bus). -r 10 Sets numeric radix for registers to 10. Default is 16. -d <level> Set the level of debugging output produced by zreg. The default level is 1.
zrld NAME zrld – ZNYX redirector daemon SYNOPSIS zrld [-d <level>] [-p <port>] [-f] DESCRIPTION zrld is used for remote management of OA/HA appl ications. OA/HA applications capable of remote management include zlc , ztats , zlmd . zrld only allows requests from hosts listed in /etc/rcZ.
zsnoopd NAME zsnoopd - IGMP Snooping daemon for the OpenArchitect switch . SYNOPSIS zsnoopd [-d <level>] [-f] [-h <hostname>] [-p <ppa>] [-r <sec>] [-t <sec>] [-u <sec.
-f Run zsnoopd in foreground. Default is to run it in background. -h <hostname> Connect to remote host <hostname>. -p <ppa> Start zsnoopd on switch <ppa>. Default is 0. -r <sec> Time to wait, in seconds, before removing a port with no router multicast traffic.
zspconfig NAME zspconfig - configure and start surviving partner SYNOPSIS zspconfig [-d <level>] [-p <directory_path>] [-u <dhcp_interface>] [-c <dhclient.conf>] [-t <timeout>] [-s] [-v] -f <file> DESCRIPTION zspconfig is used to configure and start the Surviving Partner software.
-t <timeout> Time to wait in seconds before giving up on finding a Surviving Partner to retrieve configuration information from. Only valid with the –u option. -s Do not ask for confirmation. Run from a script. -v Prints the current version of zspconfig.
A sibling_addresses statement is required for each VLAN created with t he zconfig commands. The two addresses in the list indic ate there are two switches in the Surviving Partner group. The first address 10.0.0.30 and 11.0.0.30 are assigned to the swit ch on which the configuration is being run.
monitor_only - Ports that are monitored but do not have a virtual address managed on them. They will not have their links brought down tempor arily during a failover scenario. These ports are only monitored. If a problem occ urs on this type of link it will cause a failover scenario.
interconnect, it is not a bonding driver ena bled port, and therefore should be setup as an interconnect port type. To accompl ish this, the zre ports are listed to avoid conflicting port types. Note that a single line cannot contai n both zhp and zre definitions.
Additional startup scripts may be included in the configuration using the start_script command. The files in the start_script com mand will be placed in a location for tftp transfer to sibling switches that initialize us ing the –u option.
distinguish potentially more than one backup switch. This configuration file is placed in /tftpboot, and is retrieved via DHCP by a replacement switch on boot up. /etc/rcZ.d/surviving_partner/dhcpd.conf Configuration script used by dhcpd when the switch becomes master.
zstack NAME zstack - Configures the OpenArchitect switch st acking. SYNOPSIS zstack [-h <host_name>] [-d <level>] [-a] [-t] [{-f <file>} | <configuration>] DESCRIPTION zstack combines multiple switch fabric chips into a s ingle virtual switch.
semicolon-delimited statem ents. Each statement specifies an action to ta ke on a stack. A stack is a group of ports on a single switch fabric chip. Ac tions include stack creation, stack port association, stack configuration and sta ck control. Comments, spaces and new lines are i gnored.
an equal number of ports in each list. Wild car ds may be used in the zre_lists . See below. Stack port association syntax for a Ether net Switch Blade: stack0: zre0.
are supported. enable; The enable statement turns on stac king that has been previously configured. This statement cannot be made until configuration is compl ete.
zre lists . Example of stack0..3 representing stacks 0, 1, 2 and 3. SEE ALSO zconfig Ethernet Switch Blade User's Guide release 3.2.2j page 257.
ztats NAME ztats − Display statistics and information about switch SYNOPSIS ztats [-d <level>] [-i <unit>] | [-m <port>] | [-v <vlan id>] | [-t <tgid>] | [-v] DESCRIPTION ztats displays MIB counters for a selected physic al port, trunk group or VLAN.
zsync NAME zsync − Saves changes to the flash. SYNOPSIS zsync [-c][-f][<dir_or_file>] zsync [-c][-f][-t <file>] zsync [-c][-f][-z] zsync [-c][-l] DESCRIPTION zsync is used to save a snapshot of the current file system to flash ROM.
-t <file> Read files to be saved from a tar file. -z Zero the overlay area. -f Do not confirm with user and do not warn if saving failed. Exit code can be examined to determine success or failure. <dir_or_file> Save only the named file, or save the named directory to the overlay.
ztmd NAME ztmd – traffic management daemon which accept s messages from traffic filtering and quality of service applications and se ts up hardware. SYNOPSIS ztmd [-d <level>] [-p <port>] [-f] [-i <pid>] [-o <pid>] [-a <addr>] [-l] DESCRIPTION ztmd listens for messages on a multicast port.
SEE ALSO zqosd, iptables(8), tc(8), zfilterd Ethernet Switch Blade User's Guide release 3.2.2j page 262.
brctl(8) NAME brctl - Bridge and Spanning Tree Protocol administration. SYNOPSIS brctl [options] DESCRIPTION brctl is used to set up, maintain, and display the bridge configuration in the Linux kernel. brctl is a standard command included wit h Linux bridge support which includes Rapid Spanning Tree Protocol (RSTP) support.
bridge will not send or receive BPDUs, and will thus not participate in the Spanning Tree Protocol. If your bridge isn't the only bridge on the LAN, or if there are loops in the LAN's topology, DO NOT turn this option off. Turning this option off may impair network traffic, so be careful.
SEE ALSO zconfig , zl2d Ethernet Switch Blade User's Guide release 3.2.2j page 265.
Appendix B Base Switch Com mand Man Pages OpenArchitect applications are im plemented above the OpenArchitect libraries and the R MAPI interface. OpenArchitect applicati ons are used for normal operation of the switch, for runtime status and diagnostics, and for prototyping new a pplications development.
vrrpconfig NAME vrrpconfig – Configure and control the running vrrpd SYNOPSIS vrrpconfig [-d <level>] -- <vrrpd parameters> vrrpconfig [-d <level>] [-k] [-a] [-p] [-s <vid>] DESCRIPTION vrrpconfig provides communication with a running vrrpd daemon.
EXAMPLES Here is an example of using the -- invocat ion method that changes the priority to 99 for the Virtual Router associated with the Virtual Router Identifier 1: vrrpconfig -- -v 1 –p 99 SEE ALSO vrrpd Ethernet Switch Blade User's Guide release 3.
vrrpd NAME vrrpd – Virtual Router Redundancy Protocol Daemon SYNOPSIS vrrpd -i ifname -v vrid [-f piddir] [-s] [-a auth] [-p prio] [-nhb] [-I ifname] [-d delay] [-m address] [-M ] [-B] [-S script] [-c conf_file] [-D level] ipaddr DESCRIPTION vrrpd is an implementation of Virtual Re dundant Routing Protocol (VRRPv2) as specified in RFC2338.
other words there is a unique vrid to ifname associated with the –i option. -s Toggle preemption mode (Enabled by default). Preemption means that a Master switch will go to Backup if a current Backup has higher priority. -M Become MASTER when priority is equal.
spaces. The –n option overrides the change made with –m. The result of which to use the native MAC address of the interface. Using the –n option is not recommended.
zbootcfg NAME zbootcfg − Modifies the boot parameters of t he OpenArchitect switch. SYNOPSIS zbootcfg -a | -d <device number> [<boot_string>] DESCRIPTION zbootcfg is used to display or modify the boot parame ters on the switch.
application flash. Typically this is required before updating application flash 1. B y booting the alternative image, if a fail ure occurs during the programming of application flash 1, rec overy is easier. zbootcfg -d 2 The next example passes the -i option to the booti ng kernel.
zconfig NAME zconfig - Configures the OpenArchitect switch. SYNOPSIS zconfig [-h <host_name>] [-d <level>] [-a] [-t] [{-f <file>} | <configuration>] DESCRIPTION zconfig creates VLAN groups of switch ports or trunks. Each V LAN group forms a Layer 2 switching domain.
entered as parameters to zconfig . CONFIGURATION SYNTAX zconfig takes configuration data from standar d input or from a file with the -f option. In either case, the configuration syntax is the same. The zconfig configuration data consists of a list of semicolon-delimited sta tements.
A port member is identified with the zre<X> format, where x represents a port number between 0 and 23 for the in-band ports. The Out-of-Band ports cannot be included in the List of ports. teardown Removes the trunk interface, making the ports which were part of the trunk available for configuration in other trunks or VLANs.
Network Interface Statements The syntax for a network interface state ment is the interface name followed by a colon and a n action. Network interface state ments are used to create or tear down a VLAN group and can consist of one or a list of network interfac e names; followed by a colon and then an action.
The next statement creates a VLAN group with the VID number 100 and the network interface named zhp1 . This VLAN includes four switch ports, zre1, zre10, zre11, zre13. zhp0: vlan100 = zre1,zre10,zre11,zre13; The next statement adds two switch port s, zre1, zre2 and zre3, to an existing network interfac e and VLAN.
tag are given the VLAN tag with the VID number 1, ent er: zre1=untag1; If port 0 is also a member of VLAN 100, packets f or VLAN 100 are sent from this port with a VLAN tag as part of their header. In the next example, the switch ports 10, 11, and tr unk 2 are configured as untagged members of VLAN 100.
network interface, and all ports lis ted after the equal sign are included in each group. The following statement creates 14 VLAN groups with VID numbers 1-14.
The previous configuration can be used for creat ing a 14 port Layer 2 switch, all 14 ports assigned to the same VLAN. SEE ALSO zl3d Ethernet Switch Blade User's Guide release 3.
zcos NAME zcos - class of service queue control SYNOPSIS zcos [-h <hostname>] [-d <level>] [ -u <default priority> ] [ -m q0,q1,q2,q3,q4,q5,q6,q7 ] [-n <queue length list in packe.
Each cos queue is limited in the numbe r of packets it can hold waiting scheduling; the memory used by each queue is managed to provide a guarant eed space with additional space shared among all queues for a port. OPTIONS Most options are optionally followed by a <port list>, which may include zre port ranges, like zre0.
Packets are first counted against the re served space for a queue. When that space is occupied, additional memory is used from the dynamic me mory pool until the dynamic pool usage limit f or the port is reached. Any additional packets received for the queue on this port are dropped.
-q all | qmap | qinfo | scheduler [<port list>] Queries the current COS/QOS Settings. all - Displays all of the queue mappings, queue limi ts, metering and scheduling settings qmap - Displays the priority to COS queue mappings . qinfo - Displays queue limits for the COS queues .
zdog NAME zdog - Configure and send heartbeats to watchdog e nabled drivers. SYNOPSIS zdog [-d <level>] -h | -i <interval> | -n <heartbeats> zdog [-d <level>] -b zdog [-d <l.
software component of zdog when launched from monit insures that monit is running to perform higher level tasks. And finally monit can be used to monitor any or all critical system resources and processes in the system. OPTIONS -d set debug level to <level> -h Toggle use of the hardware watchdog timer.
zffpcounter NAME zffpcounter —Query or clear one or more Fast Filter Processor (FFP) counters. SYNOPSIS zffpcounter -P <zre_port> [-p <ppa>] [-i <index>] [-h <hostname>] [-c] [-d <level>] DESCRIPTION The switch enforces filtering rules through the FFP.
EXAMPLES The first example queries all FFP counte r values. zffpcounter The output displays the initial stat e of the counters. Note that the counters are not initia lized on startup, Counter 0: 59602801 Counter 1: 83360091 Counter 2: 83361262 .
Counter 30: 0 Counter 31: 0 iptables(8) is used to setup a rule, and associate that rule with a counter. For instance, add a rule to accept all packets from 10.
Counter 19: 0 Counter 20: 0 Counter 21: 0 SEE ALSO zirule, iptables(8) Ethernet Switch Blade User's Guide release 3.2.2j page 291.
zfilterd NAME zfilterd - A daemon to use the filter hardware of the OpenArchitect switch for filtering based on iptables(8) rules. SYNOPSIS zfilterd [-d <level>] [-p <port>] [-f] [-l] [-i .
zflash NAME zflash − Loads images into the flash ROMs on the OpenArc hitect switch. SYNOPSIS zflash -d <dev> [-o|-O <offset>] <image_file> <upgradeipmi.img> DESCRIPTION zflash enables you to program the flash ROMs on the switch.
Exercise caution when using this command, a s an error can render your switch inoperable. Do not interrupt this process until complete. SEE ALSO zbootcfg Ethernet Switch Blade User's Guide release 3.
zgmrpd NAME zgmrpd - GARP Multicast Registration Protoc ol (GMRP) daemon for the OpenArchitect switch. (Partially supported in this release .) SYNOPSIS zgmrpd [-d <level>] [-f] [-h <hostname&.
Only the GARP normal registration mode is cur rently supported. Multiple instances of zgmrpd may run concurrent ly provided the targets are unique. However, zgmrpd cannot run concurrently with zsnoopd . See zsnoopd . OPTIONS -d <level> Sets the level of debugging output required by zgmrpd.
zgr NAME zl2, zl2mc, zl3host, zl3net, zvlan – Formatted display of OpenArchitect generic tables. zl2 displays the abstraction API’s layer 2 ta ble. zl2mc displays the abstraction API’s layer 2 multi cast table. zl3host displays the abstraction API’s layer 3 host route table.
Several options are available whi ch enable the user to display only selected entries. Addit ionally, there is an option that clears user -specified entries in the table. OPTIONS -i <index> Displays the entry at the <index> position in the table.
zl2 Be careful, the -c option does not ask. The foll owing command deletes all entries in the zl2 table: zl2 -c SEE ALSO zal Ethernet Switch Blade User's Guide release 3.
zgvrpd NAME zgvrpd - GARP VLAN Registration Protocol (GVRP) da emon for the OpenArchitect switch. SYNOPSIS zgvrpd [-d <level>] [-f] [-h <hostname>] [-p <ppa>] [-t <target>] DESCRIPTION zgvrpd is run after the network interfaces are created and initialized with zconfig , and started with ifconfig(1M).
default level is zero (0). Setting the debug level higher produces more output. Five (5) is currently the maximum output level. -f Run zgvrpd in foreground. Default is to run it in background. -h <hostname> Connect to remote host <hostname>.
zl2d NAME zl2d - Layer 2 daemon for the OpenArchitect switch. SYNOPSIS zl2d [start | stop] [-t <msecs>] [-d <level>] [-f] [-p <priority>] <iface..> DESCRIPTION zl2d is run after the network interfaces are created and initialized with zconfig .
OPERATIONS zl2d manages the Spanning Tree state fi elds in the switch of each port within the bridge(s). Based on a timer, zl2d reads the port inform ation for each Linux bridge and updates the switch when necessary.
zl3d NAME zl3d - Layer 3 daemon for the OpenArchitect switch. SYNOPSIS zl3d [-h <host_name>] [-t <msecs>] [-b] [-e] [-l] [-n] [-d <level>] <iface ..> DESCRIPTION zl3d is run after the network interfaces are created and initialized with zconfig , and started with ifconfig(1M).
produces more output. Four (4) is currently the maximum output level. <iface…> The network interfaces on which zl3d should operate. These network interfaces must first be created by zconfig. zl3d does not operate with standard network interface cards.
zlc NAME zlc − link and LED control SYNOPSIS zlc [-h <hostname>][-d <level>][-x] <port_list> <action> [on | off ] zlc [-h <hostname>][-d <level>][-x] <action&g.
must be down to change the speed. Assuming zre1 i s part of interface zhp0 , ifconfig zhp0 down zlc zre1 100fd The external fault, internal fault , and ok LEDs can be set on a per port basis or globally .
zlmd NAME zlmd − monitor link changes or hot swap events. SYNOPSIS zlmd [-h <hostname>] [-b] [-d <level>] {-f <file>} | <configuration> DESCRIPTION The zlmd application is intended to run as a daem on, waiting for a configured event to occur and then running the program configured for that event.
<port-list> A list of ports in the same forms supported by zconfig, e.g. zre1,zre2 or zre10..14 <program> Path to an executable program or script to be run when the event occurs.
zlogrotate NAME zlogrotate − Rotates log files. SYNOPSIS zlogrotate [-b] [-t time] [-s segment size] [-n # of files] [-f file to rotate] DESCRIPTION zlogrotate rotates the selected file every [t ime] seconds if the file is larger than [se gment size].
zmirror NAME zmirror - Set packet mirroring on an ingress or egress port SYNOPSIS zmirror -a | -t zmirror [-e] <from_list> <to_port> DESCRIPTION zmirror sets packet mirroring from a given set of ports to one given port . Turning on packet mirroring causes a copy of the packet to be se nt to the to port.
zmirror zre1, zre2, zre3 cpu Is the same as: zmirror zre1 cpu zmirror zre2 cpu zmirror zre3 cpu Multiple mirroring setups can be ma de. The following example will mirror port 1 traffic to port 11 and port 2 traffic to port 12. zmirror zre1 zre10 zmirror zre2 zre11 Setting a different to port will ove rwrite the previous setting.
, (comma) Use for creating lists .. (dot-dot) Specifies an inclusive range Below are some examples for the correct usage of the comma (,) and dot-dot (..). Each line below produces the same results: zre1, zre2, zre3, zre4 zre1..4 zre1, zre2..4 zre1..2, zre3.
zmnt NAME zmnt − Expands the read/write files onto the RAM dis k. SYNOPSIS zmnt [-c] <directory> zmnt [-c] -t <file> zmnt [-c] –l DESCRIPTION zmnt expands files from flash onto the RAM disk that have been previ ously saved with zsync .
zmnt –t overlay.tar The resulting tar file can now be saved on a dif ferent host as a snapshot of the overlay at that point in time. Use zsync to restore the ove rlay on the switch: zsync –t overlay.tar The restored overlay will be used upon the ne xt reboot.
zpeer NAME zpeer – Application for High Availability comm unication between the Fabric and Data switches. SYNOPSIS zpeer [-d <level>] local|peer <command> <value>|query zpeer [-d &.
displayed by query, but must be set at initial ization. After setting the healthy state, the query will retur n the backup state. backup Used to reflect the backup state of vrrpd master Used to reflect the master st ate of vrrpd The priority value is a value betwee n 0 and 255.
be also reset. SEE ALSO zspconfig Ethernet Switch Blade User's Guide release 3.2.2j page 318.
zqosd NAME zqosd – monitors tc(8) commands to implement classification filters and queuing disci plines in hardware. SYNOPSIS zqosd [-d <level>] [-p <port>] [-f] [-l] [-i <pid>] [-.
SEE ALSO ztmd, tc(8), zfilterd Ethernet Switch Blade User's Guide release 3.2.2j page 320.
zrc NAME zrc - Packet rate control SYNOPSIS zrc -b | -m | -d | -t | -a [-p <port>] [-v <vlan>] [-g <group>] [-M <mac_addr>] [-T <timeout>] [-D <level>] <rate> DESCRIPTION zrc sets rate control on Broadcast, Multicast and/or Destination Lookup Failure (DLF) packets.
zreg NAME zreg - Read and write registers and tables on the OpenArchite ct switch switching hardware. SYNOPSIS zreg [-p <ppa>] [-w] [-i <index>] [-t <index>] [-k] [-h <hostname>.
connected (i.e., the one that is on the local PCI bus). -r 10 Sets numeric radix for registers to 10. Default is 16. -d <level> Set the level of debugging output produced by zreg. The default level is 1. Setting the debug level higher produces more output.
zrld NAME zrld – ZNYX redirector daemon SYNOPSIS zrld [-d <level>] [-p <port>] [-f] DESCRIPTION zrld is used for remote management of OA/HA appl ications. OA/HA applications capable of remote management include zlc , ztats , zlmd . zrld only allows requests from hosts listed in /etc/rcZ.
zsnoopd NAME zsnoopd - IGMP Snooping daemon for the OpenArchitect switch . SYNOPSIS zsnoopd [-d <level>] [-f] [-h <hostname>] [-p <ppa>] [-r <sec>] [-t <sec>] [-u <sec.
-f Run zsnoopd in foreground. Default is to run it in background. -h <hostname> Connect to remote host <hostname>. -p <ppa> Start zsnoopd on switch <ppa>. Default is 0. -r <sec> Time to wait, in seconds, before removing a port with no router multicast traffic.
zpeer peer state query The output from the above command during the boot process would be “unhealthy” The -a option can be used to display a complete listing of all state and priority information and internal information that can be use d for debugging.
zspconfig NAME zspconfig - configure and start surviving partner SYNOPSIS zspconfig [-d <level>] [-p <directory_path>] [-u <dhcp_interface>] [-c <dhclient.conf>] [-t <timeout>] [-s] [-v] -f <file> DESCRIPTION zspconfig is used to configure and start the Surviving Partner soft ware.
finding a Surviving Partner to retrieve configuration information from. Only valid with the –u option. -s Do not ask for confirmation. Run from a script. -v Prints the current version of zspconfig. -f <file> The provided <file> is used as input to configure the Surviving Partner.
A sibling_addresses statement is required for each VLAN c reated with the zconfig commands. The two addresses in the list indicate there are two switches in the Surviving Pa rtner group. The first address 10.0.0.30 and 11.0.0.30 are assigned to the sw itch on which the configuration is being run.
monitor_only - Ports that are monitored but do not have a virtual address managed on them. They will not have their links brought down tempor arily during a failover scenario. These ports are only monitored. If a problem occ urs on this type of link it will cause a failover scenario.
interconnect, it is not a bonding driver ena bled port, and therefore should be setup as an interconnect port type. To accompl ish this, the zre ports are listed to avoid conflicting port types. Note that a single line cannot contai n both zhp and zre definitions.
Coordination between the data and fabr ic switches can be enabled by setting the board_synchronization_mode . Possible modes are “off” and “basic”. B oard synchronization is off by default. When board sync hronization is put into basic mode, HA events on the base switch are coordinated with the HA events on the fabric switch.
files run the Surviving Partner. This is a li st of all configuration and script files: /etc/rcZ.d/S70Surviving_partner The main startup script that starts the Surviving Partner by running zconfig, ifconfig, zlmd and vrrpd. zspconfig prompts the user to run this script.
Ethernet Switch Blade User's Guide release 3.2.2j page 335.
zstack NAME zstack - Configures the OpenArchitect switch st acking. SYNOPSIS zstack [-h <host_name>] [-d <level>] [-a] [-t] [{-f <file>} | <configuration>] DESCRIPTION zstack combines multiple switch fabric chips into a s ingle virtual switch.
semicolon-delimited statem ents. Each statement specifies an action to ta ke on a stack. A stack is a group of ports on a single switch fabric chip. Ac tions include stack creation , stack port association , stack configuration and stack control . Comments, spaces and new lines are i gnored.
semicolon are two zre_lists separated by an equal si gn. The first is the list of virtual port names, the second is the physical port names. The assignment is done in order, and there must be an equal number of ports in each list. Wild car ds may be used in the zre_lists .
Stack Control Statements Finally after creating the sta ck, associating the ports, and setting the stack confi guration, the stack can be enabled using one of the Stack Cont rol statements. The following stack control state ments are supported. enable; The enable statement turns on stac king that has been previously configured.
ztats NAME ztats − Display statistics and information about switch SYNOPSIS ztats [-d <level>] [-i <unit>] | [-m <port>] | [-v <vlan id>] | [-t <tgid>] | [-v] DESCRIPTION ztats displays MIB counters for a selected physic al port, trunk group or VLAN.
zsync NAME zsync − Saves changes to the flash. SYNOPSIS zsync [-c][-f][<dir_or_file>] zsync [-c][-f][-t <file>] zsync [-c][-f][-z] zsync [-c][-l] DESCRIPTION zsync is used to save a snapshot of the current file system to flash ROM.
OPTIONS -c Save files to the custom overlay -t <file> Read files to be saved from a tar file. -z Zero the overlay area. -f Do not confirm with user and do not warn if saving failed. Exit code can be examined to determine success or failure. <dir_or_file> Save only the named file, or save the named directory to the overlay.
ztmd NAME ztmd – traffic management daemon which acc epts messages from traffic filtering and qua lity of service applications and sets up hardwa re. SYNOPSIS ztmd [-d <level>] [-p <port>] [-f] [-i <pid>] [-o <pid>] [-a <addr>] [-l] DESCRIPTION ztmd listens for messages on a multicast port.
SEE ALSO zqosd, iptables(8), tc(8), zfilterd Ethernet Switch Blade User's Guide release 3.2.2j page 344.
brctl(8) NAME brctl - Bridge and Spanning Tree Protocol administration. SYNOPSIS brctl [options] DESCRIPTION brctl is used to set up, maintain, and display the bridge configuration in the Linux kernel. brctl is a standard command included wit h Linux bridge support including Rapid Spanning Tree Protocol (RSTP).
makes the interface a port of the bridge. This means that all frames received on the interface will be processed as if destined for the bridge. Also, when sending frames on the bridge, the interface will be considered as a potential output interface.
setmaxage <bridge> <time> sets the bridge's maximum message age to <time> seconds. setpathcost <bridge> <port> <cost> sets the port cost of the port to <cost>. This is a dimensionless metric. The path cost is set to 100 for all OpenArchitect switch ports by default.
Appendix C Intellig ent Platform Management Interface The Ethernet Switch Blade provides Inte lligent Platform Management Interface (I PMI) support. IPMI circuitry provides: • The communication channel between the Base board Management Controller (BMC) and the CPU for management.
M States M5 Deactivation r equest (e.g. hot swap ejector opened) M6 Deactivation gr anted by ShMC M7 Unexpected los s of communication between FRU and ShMC Table C.1.: IPMI M States Peripheral Management Controller Functi onal Support The following IPMI commands are implemented in vers ion 1.
Sensor Reading Example This is an example of how to structure a c ommand and response to determine a sensor value. In this example, a GetSensorReading com mand is sent from BMC (address 20h in this example), to the switch in slot 2 (geographical addre ss is B2h) to read the temperature sensor, which is assigned to sensor number 60h.
Standard IPMI Response: GetSensorReading Byte Data Field D escription 1 rqAddr 20h 2 netFn/Lun 16h 3 check1 CAh 4 rsAddr B2h 5 seq no 06 6 command 2Dh 7 completion code 00h 8 sensor reading 1Bh -> 27 Celsius degree 9 optional data byte C0h scanning is enabled 10 optional data byte C0h 11 optional data byte 00 12 checksum2 80h Table C.
Structure of Standard IPMI Commands: From BMC to PMC Structure of Standard IPMI Commands BMC - PMC Byte Data Field D escription 1 rsAddr <slot’s IPMB addr> 2 netFn/Lun <netFn> 3 check1 &.
Event Generator The PMC’s event generator is disabled until it receives a SetEventReceiver command from BMC for Event Receiver’s slave address and LUN. If the event generator is enabled, PMC reports significant events to the BMC a synchronously. The standard IPMI platform event message format is used.
Version 2.x supports three FRU Inventory Device Com mands: Get FRU Inventory Area Info Read FRU Data Write FRU Data The spare SEEPROM space is allocated as fol lows: Spare SEEPROM Space Allocation Spa.
IPMB Override Status Data 5= Clock low time out; 6=under test; 7=Undiagnosed communications failure Table C.9.: IPMB Override Status Data Ethernet Switch Blade User's Guide release 3.
Index Index adduser........................................................................................................................................................................................... ............................................
Index Network Time Protocol..................................................................................................................................................... 72, 131 NFS...............................................................
Index tc 62, 113 The U32 Filter................................................................................................................................................................... 66, 124 thttpd..........................................
Index ztmd......................................................................................................................................................................................... 301 zvlan..............................................
Un point important après l'achat de l'appareil (ou même avant l'achat) est de lire le manuel d'utilisation. Nous devons le faire pour quelques raisons simples:
Si vous n'avez pas encore acheté ZNYX bh5700 c'est un bon moment pour vous familiariser avec les données de base sur le produit. Consulter d'abord les pages initiales du manuel d'utilisation, que vous trouverez ci-dessus. Vous devriez y trouver les données techniques les plus importants du ZNYX bh5700 - de cette manière, vous pouvez vérifier si l'équipement répond à vos besoins. Explorant les pages suivantes du manuel d'utilisation ZNYX bh5700, vous apprendrez toutes les caractéristiques du produit et des informations sur son fonctionnement. Les informations sur le ZNYX bh5700 va certainement vous aider à prendre une décision concernant l'achat.
Dans une situation où vous avez déjà le ZNYX bh5700, mais vous avez pas encore lu le manuel d'utilisation, vous devez le faire pour les raisons décrites ci-dessus,. Vous saurez alors si vous avez correctement utilisé les fonctions disponibles, et si vous avez commis des erreurs qui peuvent réduire la durée de vie du ZNYX bh5700.
Cependant, l'un des rôles les plus importants pour l'utilisateur joués par les manuels d'utilisateur est d'aider à résoudre les problèmes concernant le ZNYX bh5700. Presque toujours, vous y trouverez Troubleshooting, soit les pannes et les défaillances les plus fréquentes de l'apparei ZNYX bh5700 ainsi que les instructions sur la façon de les résoudre. Même si vous ne parvenez pas à résoudre le problème, le manuel d‘utilisation va vous montrer le chemin d'une nouvelle procédure – le contact avec le centre de service à la clientèle ou le service le plus proche.
