Help VMware esxi vm High CPU running SharePoint server 2019

Occasional Contributor

Environment: VMware esxi vm running SharePoint server 2019 on Server 2019.

Issue: We had to reboot the server while working another issue. Since about that time the VM CPU usage has been way up. It was averaging 20ish % before reboot and now is at 70-80%.

 

We have rebooted the server for other maintenance this year without a problem.  The only thing I can think of is that SQL server VM got cloned, so 2 were in production at same time.  The clone has since been shutdown.  Could SharePoint be running extra services or processes looking for cloned vm? 

 

I'm new to SharePoint server so I'm new share what is 'normal'.  The VM has 10gig memory allotted.

jcmfxt_0-1654091391667.png

jcmfxt_1-1654091459809.png

 

Thank you for the help.

12 Replies

@jcmfxt your issue requires bit more info:

 

  • Which is the SharePoint Architecture topology ? Single server , Multiple Servers etc..
  • Which MinRole has been assigned to the impacted server ?

The problem with SQL may be related.

It depends how you have configured the instance at the very beginning in SP:

If you used alias name than it should not be a problem, meaning you just swap the instance behind the alias at the moment of the SQL server clone and the rest is transparent for SharePoint.

If not than you have to check each SP service (MMS, Search, BDC etc...) and see where they are pointing to

Thanks for replying. I'm new to SharePoint and this company and everything was already deployed. We have a single server with only about 15 users. In Central admin is says role is custom. I'm not sure where to check if there's an alias or where the SP services are pointing.

@jcmfxt 

 

Custom role means you are not taking benefits of the MinRole feature and you can decide every SP service startup on your own (as we are talking of a single server it's fine).

 

For checking the alias click on the server start and type "cliconfg.exe", the SQL Server Client Network Utility should pop up and you have the Alias tab to check whether an alias is present or not

 

You can match later the alias information with the Farm Information that you can find in the Servers in Farm tab in the Central Admin (/_admin/FarmServers.aspx)

 

To check the DB endpoint configured for the SP services, one option is to use the CA, click on "Manage Service Applications", select the SP Service then click on Properties. Another option is to use PowerShell

@mr_w1nst0n 

 

I ran the cliconfg and there wasn't any aliases.  I went to service applications and there's not an SP listed.

jcmfxt_0-1654189350896.png

 

@jcmfxt 

the list you provided shows all the SP services configured in your farm (and relative proxy)

 

You have to click on each SP service (eg. App Management Service) and then select Properties on the top bar navigation. Do it only for the service (1st line) and not for the proxy (2nd line)

 

A pop up should appear showing the configuration of the service itself

From there you should be able to see the name of the Database and the SQL Instance.

 

Check if there is any discrepancies based on what you mention about the SQL Instance clone.

 

Side Note: You can also check if any SP major issue by analyzing the ULS logs. There is a diagnostic tool provided by MSFT for that: View diagnostic logs in SharePoint Server - SharePoint Server | Microsoft Docs

I wish I knew what I was looking at .. I've read somewhere to look at the Correlation ID but the script never finishes.
eg: Get-SPLogEvent | ?{$_.Correlation -eq "6922c76b-32f7-4e2e-af96-fbda88a77fa5"}

I did use PowerShell to view to 'critical' errors. Here's a few of the errors:
-w3wp.exe: All query processing components are in 'Failed' status.
-Unknown SQL Exception 18452 occurred. Additional error information from SQL Server is included below. Login failed. The login is from an untrusted domain and cannot be used with Windows authentication.
-UsageAnalyticsTimerJob-xxxxxxxxxxxxxxxxx : Failed to retrieve Analysis Engine Service.
-The Execute method of job definition Microsoft.Office.Server.UserProfiles.FeedCacheRepopulationJob (ID xxxxxxxxxxxxxxxxxx) threw an exception. More information is included below. There are no addresses available for this application.. (Correlation=xxxxxxxxxxxxxxxxxxx)

@jcmfxt 

 

It seems you have multiple issues:

  1. The Search Service seems to be in error state (w3wp error message)
  2. You have login issues in SQL instance (you have to debug using SQL logs to see which is the exact account/s generating issues and where it is used in SharePoint)
  3. The UsageAnalyticsTimerJob is correlated with the Search Service error state
  4. The User Profile service error could be caused by the Distributed Cache Service. The log does not says much, you have to check deeper in the ULS logs 

@mr_w1nst0n 

I was reading on another forum to reset Index?

Search Service Application

jcmfxt_0-1654605666604.png

 

@jcmfxt 

Yes this fix the issue usually 90% of time.

Remember that when you perform the index reset, no search result will appear for some time, SP takes time to index again all contents.

Depending on the volume of documents you have, you may schedule it during the week-end / outside business hours

Well.. I went to Reset Index and got an error. Then I noticed in CA an link to 'Resume/Pause' so I clicked resume and got error "There was no endpoint listening at net.tcp://ServerNameHere/2B8A34/AdminComponent1/Management/NodeController that could accept the message. This is often caused by an incorrect address or SOAP action"

@jcmfxt 

Log into the SharePoint server

Open the SharePoint PowerShell Management Console

Try to Resume the search via PowerShell

$ssa = Get-SPEnterpriseSearchServiceApplication
Resume-SPEnterpriseSearchServiceApplication $ssa