Performace of HBA

[sjadapa]

Hi

i am running ESX 3.5 on IBM z series blade server. It has got two HBA's. Now only one HBA is used, as the other is used as failover.

Can any body help in finding the performance of the used HBA.

I want to know the I/O traffic on the HBA.

Any inputs and suggestions will be highly appreciated.

****If you find this or any other answer useful please consider awarding points by marking the answer CORRECT or HELPFUL ****

Jadapa RHCE, MCSA

[sjadapa]

any suggestion...

How to check the utilizaiton of HBA..

[mcowger]

You can't at the HBA level (should your switch might tell you)

YOu should check it at teh LUN level, which is exposed in the DISK performance counters.

--Matt

VCP, vExpert, Unix Geek

[sjadapa]

Matt,

Can sansurfer be used to check the hba peformance, as i am using qlogic HBA's

Is there any other way to check the utilization of HBA

Thanks

[mcowger]

I gave you my best ways - if you want others you need to do your own research.

--Matt

VCP, vExpert, Unix Geek

[srinivasvivek]

Won't ESX top provides the info?

Run esxtop from console and press 'd' will give HBA level stats

Please mark if helpful

[sjadapa]

Thanks matt,

Can you provide some guidence on to how to check at LUN level.

[srinivasvivek]

http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=100311...

Check this out

[sjadapa]

Hi Srinivas,

These are the outputs of HBA storage performance from esxtop. can you tell me what the below matrix mean...

ADAPTR CID TID LID WID NCHNS NTGTS NLUNS NVMS SHARES BLKSZ AQLEN LQLEN WQLEN

vmhba1 - - - - 1 1 1 6 - - 127 0 0

vmhba2 - - - - 1 3 45 443 - - 4096 0 0

vmhba3 - - - - 1 3 45 443 - - 4096 0 0

[srinivasvivek]

Here you go..

ADAPTR The name of the storage adapter.

CID The storage adapter channel id. This id is only visible if the corresponding adapter is expanded. See interactive command 'e' below.

TID The storage adapter channel target id. This id is only visible if the corresponding adapter and channel are expanded. See interactive commands 'e' and 'a' below.

LID The storage adapter channel target LUN id. This id is only visible if the corresponding adapter, channel and target are expanded. See interactive commands 'e', 'a' and 't' below.

WID The storage adapter channel target LUN World id. This id is only visible if the corresponding adapter, channel, target and LUN are expanded. See interactive commands 'e', 'a', 't' and 'l' below.

NCHNS The number of channels.

NTGTS The number of targets.

NLUNS The number of LUNs.

SHARES The number of shares.

BLKSZ The block size in bytes. This statistic is only applicable to LUNs.

AQLEN The storage adapter queue depth.

LQLEN The LUN queue depth.

WQLEN The World queue depth.

ACTV The number of commands in the ESX Server VMKernel that are currently active.

QUED The number of commands in the ESX Server

CMDS/s The number of commands issued per second.

READS/s The number of read commands issued per second.

WRITES/s The number of write commands issued per second.

MBREAD/s The megabytes read per second.

MBWRTN/s The megabytes written per second.

DAVG/cmd The average device latency (millisecs) per command.

KAVG/cmd The average ESX Server VMKernel latency (millisecs) per command.

GAVG/cmd The average Guest OS latency (millisecs) per command.

ABRTS/s The number of commands aborted per second.

RESETS/s The number of commands reset per second.

[sjadapa]

I got the info srinivas.

Is there any way to check the graphical

[srinivasvivek]

You need to use vCenter and go to the performance tab. Select disk or network based on what you need

Please mark if answer is useful

[sjadapa]

Hi Friends,

Qlogic support has answered my issue.

Thanks srinivas and matt for the support

Jadapa RHCE, MCSA

[bulletprooffool]

Contact the HBA provider - tools like powerpath etc will show you everything you need.

Alternatively, see if you can find SNMP MIBs for the counter sthat you would like.

[sjadapa]

Can you provide some more info on snmp mibs...

[mnasir]

,

You can access HBA performance monitor by using esxtop (d) command from the shell,

esxtop (press enter)

d (press enter0

Please see attached image.

Please award point if you find this post helpfull.

[mnasir]

Did you get a chance to test what I recommended on my earlier post? There are other tools out there, that can help you monitor HBA performance on real-time. But I don't usually recommend my clients using that – unless it is absolutely required by your IT policy -- here are some of the reasons why I don’t think monitoring HBA will help you that much…

HBAs are rarely the bottle neck, you will max out your LUN, or your FC fabrics before you saturate your HBAs. Also, how you design your SAN has a direct correlation to your performance than your HBAs. As a rule of thumb, try to use more spindles per array, also if you have options to use MRU vs Active/Active, obviously I would go for active/Active configuration.

Esxtop (d) in combination vCenter disk performance graph should be more than enough to monitor the status of the HBA performance - usually there are no need to use an external agent to monitor your HBA traffic, I usually try to keep esx host as clean as possible - no external agent. I treat it more like an appliance.

Please award points if you find this post helpful.

[sjadapa]

Nasir,

I followed the suggestion. I was using esxtop to find out the performance of HBA's.

Is there any other tools which can be help for to mornitor HBA for some time.

****If you find this or any other answer useful please consider awarding points by marking the answer CORRECT or HELPFUL ****

Jadapa RHCE, MCSA

[mnasir]

Try the link below - you should be able to use the Linux version.

http://www.manageengine.com/products/opstor/hba-management.html

Thanks,

Meraz

[sjadapa]

Any opensource tools...

****If you find this or any other answer useful please consider awarding points by marking the answer CORRECT or HELPFUL ****

Jadapa RHCE, MCSA