Manuel d'utilisation / d'entretien du produit BL465C G7 du fabricant HP (Hewlett-Packard)
Aller à la page of 30
Performance of HP ProLiant BL465c G 7 with AMD Opteron 6100 Series processors in 32 - and 64 -bit HP SBC environments Technical white paper Table of content s Executive summary ..........................................................................
2 Executive summ ar y Designed for virtualization yet flexible enough for an y business applicati on, t he HP ProLiant BL465c G7 server blade combines best- in -class features to help reduce netwo rk sprawl and lo wer infrastructu re costs.
3 Introduction Figure 1 sho ws the HP ProLiant BL465c G7 s erver blade. Figure 1 . The H P ProLiant BL465c G7 server bl a de, illustrating a un ique dra wer design th at allows y ou to service h ard d.
4 HP Integrated Lights -Out 3 (iLO 3) Serial Attached SCSI (SAS) Thermal logic The following secti ons of this paper desc ribe testing per formed by HP to characteri ze the performance and scalability of an HP ProLiant BL465c G7 serv er blade in 32 - and 64 -bit HP SBC environments .
5 Table 1 . Compon ents of TSScaling Component Description Automation tool s Robosrv.exe Drives the server-s ide of the loa d simulation Robocli.exe Helps drive the client-side of th e load simulation Test tools Qidle.exe Determines if any scripts have f ai led and re quire operator interventi on Tbscript.
6 Test scenario To characteri ze its sc alability, HP tested the HP ProLiant BL465c G7 server blade in 64 - and 32 -bit HP SBC environments wh en runni ng a workload based on the User pro file described above. Note: As recommend ed by HP for m ost applications, the FBWC was conf igured with 100% write for all testing.
7 Test topology Figure 2 illu strates the HP SBC test environm ent. Figure 2 . The tested envir onment (6 4-bit implementation sh own) Note: Test environmen ts such as that shown in Figure 2 are available to cus tomers at HP Solution Centers to help solv e a wide v ariety of bus iness probl ems.
8 Configurations This section ou tlines the configurations of servers and clients u sed in the test environm ent. Table 3 presents inform ation on system c onfigurations; Table 4 provides a sy stem su mmary.
9 Table 3 . Syst em configurat ions (continued) Server Configuration Domain controller 2P HP ProLiant DL3 60 G5 s erver with: Dual-core Intel Xeon p rocessor (3 .
10 Table 4 . Syst em summary for the bare-metal HP ProLiant BL465c G7 server bl a de (continued) Component Description Boot device DeviceHardd iskVolume1 Locale United States Hardware abstract i on layer 5.2.3790.3959 (srv0 3_sp2_rtm .0702 16- 1710) / 5.
11 The following sc enarios were tested: Server blade conf igured with the AMD Opter on process or Model 6174 , 6164 HE, 6136 , and 6128 HE HP SBC test environment s – 64 -bit: 6174, 616 4 H.
12 Figure 4 sho ws normalized Perfm on values. Figure 4 . Ther e were no st opped sessions when 48 7 Heavy Users wer e logged on With 487 users ac tive, there were no sto pped sessions. The processor became saturated with approximately 5 00 users lo gged on, which coincided with th e marked increase i n processor qu eue length shown in Figure 3.
13 Canary times Figure 5 sho ws sample results fo r the tested server when ru nning a typical canary script. Indi vidual user response times a re shown in blu e, with a yellow line depi cting average res ponse times.
14 Figure 6 . % Pr ocessor Time values – show ing that 141 Heavy U sers were su p ported when system PTEs became exhausted The optimal nu mber of Heavy Users support ed by the HP ProLiant BL465c G7 server blade was 141 .
15 Figure 7 . % Pr ocessor Time values – show ing that 447 Heavy U sers had been able to log on when process or utilization reached 80% Thus, as many as 447 Heav y Users were supported when pro cessor utilizati on reached 80%. Figure 8 sho ws normalized Perfm on values.
16 Figure 8 . Although 447 Heavy Users were able to log on , tw o session s had stopped With 447 Heavy User s – the number of users logged on when processor utilization reached 80% – there were two stopped sessions. Thus the opti mal number of users is (447 – 2); that is, 445.
17 Canary times Figure 9 sho ws sample results fo r the tested server when running a typical canary script. Individual user response times are shown i n blue, with a yellow line depicting average res ponse times.
18 Figure 10 . % Process or Time values – sho wing that 411 Heavy Users were supported wh en process or utilization reached 80 % The optimal nu mber o f Heavy Users supported by the bare-metal HP ProLiant BL465c G7 ser ver blade was 411 . Canary times Figure 11 sho ws sample resu lts for the tested server when ru nning a typical canary sc ript.
19 Figure 11 . Canary time v alues show that user respon se times starte d to become un acceptable when 475 Heavy Users were active Response times wer e acceptable when 411 Heavy User s – the optimal number – were active. 32 -bit test environment HP ran a perfo rmance test using a wo rkload based on the Heavy User profile.
20 Figure 12 . % Processor Tim e values – showing that 161 Heavy User s were supporte d when disk utilizat ion began to incre ase exponentially Disk utilization increased exponentially du ring this test ru n due to lack of system PTEs. The optimal nu mber of users suppo rted by this HP ProLian t BL465c G7 server blade was 161.
21 Figure 13 . % Process or Time values – sho wing that 380 Heavy Us ers were sup p orted when pr ocessor utilization reach ed 80% The optimal nu mber of Heavy Users support ed by the bare -metal HP ProLiant BL465c G7 ser ver blade was 380 .
22 Figure 14 . Canary time v alues show that user respon se times starte d to become un acceptable when 406 Heavy Users were active Response times wer e acceptable when 380 Heavy Users – the optimal number – were active. Test analysis su mmary Figure 15 su mmarizes the test results.
23 Important: As with any labor a tory benchmar k, the perf ormance metr ics quoted in this performance bri ef are ideali zed. In a production env i ronment, th ese metr ics may be impacted by a variety of factors; for more information, ref e r to Appendix B – SBC solution sizing .
24 Recommendations This section pro vides general r ecommenda tions for im proving the performance o f HP SBC servers. x64 platforms Since x64 platforms allo w you to better utilize memory an d multi -core processors, the bottleneck you are most likely to encounter 8 would be assoc iated with the di sk subsystem.
25 Note: HP has not yet ch aracterized FBWC per formance in the HP SBC environment. Howe ver, testing perf ormed using BBWC demonstr ates that , typically, improv ements from write cache are most sign.
26 Note: For more inform ation on be st practi ces associate d with virtuali zation, r efer to the HP white paper, “ Best pr actices for depl oying Citrix X enApp on XenServer .” Conclusions When planning an x64 HP SBC environment, you should select servers equipped with multi -c ore processors to help maximize scalabili ty.
27 Appendix A – U sing Microsoft Window s Serve r 2003 x64 Editions Microsoft off ers operating systems that support both 64 - and 32-bit applicati ons, as well as existing 32 -bit deployment and management too ls – all on the same platform.
28 Appendix B – SBC soluti on sizing As with any laborato ry benchm ark, the perform ance metrics quoted in this performance brief are idealized. In a production environment, these metrics may be i .
29 Figure B-1 . Home page for th e HP Citrix XenApp an d Microsoft T erminal Serv ices Sizing Tool Based on informati on provided by the cu stomer, the sizer can provide a quick, consistent mechanism for identifying the “best - fit” server f or a par ticular HP SBC environment and generat e a Bill of Materials (BOM) f or that server.
For more inform ation HP ProLiant BL465c G7 server blade http://www.hp.co m/servers/bl465c HP ActiveAnswers sol ution area f or HP SBC , including C itrix XenApp and Microsof t Terminal Services http://www.
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 HP (Hewlett-Packard) BL465C G7 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 HP (Hewlett-Packard) BL465C G7 - de cette manière, vous pouvez vérifier si l'équipement répond à vos besoins. Explorant les pages suivantes du manuel d'utilisation HP (Hewlett-Packard) BL465C G7, vous apprendrez toutes les caractéristiques du produit et des informations sur son fonctionnement. Les informations sur le HP (Hewlett-Packard) BL465C G7 va certainement vous aider à prendre une décision concernant l'achat.
Dans une situation où vous avez déjà le HP (Hewlett-Packard) BL465C G7, 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 HP (Hewlett-Packard) BL465C G7.
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 HP (Hewlett-Packard) BL465C G7. Presque toujours, vous y trouverez Troubleshooting, soit les pannes et les défaillances les plus fréquentes de l'apparei HP (Hewlett-Packard) BL465C G7 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.
