ESX Performance Monitoring

I have a SQL job that can run in mid-day and perform in a short period (1 hour).  I have the same job run during peak hours and get 2.3 hours until completion.  I have viewed storage and see no bottleneck.  Is there a type of monitoring that I can use to monitor ESX performance during this job so that I can see where I'm maxing out at?
Who is Participating?

[Webinar] Streamline your web hosting managementRegister Today

SunshineVKConnect With a Mentor Commented:
In that case you can invoke esxtop in batch mode, something like this should do

"esxtop -b /<folderpath>/esxtopdata.csv"

You can then import this into an excel workbork and manipulate it for the required graphs, charts etc etc.
but just a word of caution, depending the data collection interval, the log file can be huge.

You can also use the esxplot tool developed by Geoff White as GUI for all the data tha you collect
SteveIT ManagerCommented:

Try for free monitoring tools for virtual servers as well as other useful tools
vmwarun - ArunCommented:
Are you using vCenter or Virtual Center to manage your ESX/ESXi Hosts ?
Upgrade your Question Security!

Your question, your audience. Choose who sees your identity—and your question—with question security.

I might suggest starting inside the sql virtual machine with the normal tools (taskmgr and perfmon) to see if the vm is starved for cpu, disk, or network.  Once you identify that, then look at the tools above to identify which vm's are contending for that particular resource during your bottleneck.
use the performance tab in vcenter. from there you can view realtime (20secs), a day, week, month, etc.  You want to monitor the big four,  CPU, memory, disk , and network,

Example check
CPU ready should be 0-5%
over 10% performance issue

memory: check balooning

securitythreatAuthor Commented:
All of the above have been done.  The only tool that will help verify is from the ESX side and not the guest.  I have done the guest and see the issue.  However, know I have to determine what the root cause is.

You could try to run esxtop on the ESX server during peak and offpeak time and check the counters.

The following link might help in understanding the parameters

Check the %WAIT for vmkernel resources as well as the disk throughput stats and queue stats.
securitythreatAuthor Commented:
Sunshine, you are right on it.  However, I need to chart that data for presentation like purposes to present to team for performance review.
vmwarun - ArunCommented:
Try to use ManageEngine OpManager tool, setup the counters at both ESX and VM level since OpManager has got both Server Performance and Application Performance monitoring capabilities to monitor for a week.
Once monitoring has been done, you can use the charts to check the root cause.
securitythreatAuthor Commented:
Data is not complete but partial answer.
All Courses

From novice to tech pro — start learning today.