Highlighted
Absent Member.
Absent Member.
1982 views

HPSM disappears

So this morning, around 11:15am EDT, reports from HP Service Manager users here locally began coming in saying they were no longer able to access the server through the web server.

 

We're using 9.32.

 

When we tried our web clients, we too got back errors from IE saying that no pages were available.

 

We logged into the server itself, and the services were running - and in fact, command line invocations of sm -reportstatus were working.

 

We found no errors in the log files.

 

The HPSM service was restarted - but that did not help.

We restarted the Tomcat service - but that did not help.

We rebooted the server - and after a long wait (10 minutes or so), things began working again.

 

Nothing has been found in any log file that would indicate the existence of an issue.

 

Has anyone else encountered this type of thing?  If so, have you found a way to identify the issue or to work around it?

 

Thank you

0 Likes
31 Replies
Highlighted
Micro Focus Expert
Micro Focus Expert

When troubleshooting did you try bypassing the webserver (i.e. port 80) and seeing if you could access the Service Manager login page hitting Apache Tomcat directly using port 8080?

Highlighted
Absent Member.
Absent Member.

We didn't think to do that - or to try the native client to see if it could access things.

 

Hopefully there's never a chance to do these again - but if there is, we will try some additional things.

 

This being production, we were in a bit of a hurray to get things back up rather than experiment ...

0 Likes
Highlighted
Micro Focus Expert
Micro Focus Expert

True. I can relate to that. If it does happen again try 8080 and if it works then you've narrowed to down to - at least - one possible asset.

Highlighted
Absent Member.
Absent Member.

That's a good suggestion.

 

Since there were no indications in the logs of a problem, and the server, database, and vm admins, as well as the storage admins, all have indicated they had nothing in their logs, what could be done now in terms of monitoring and logging so that the next time HPSM just stops responding to users there is something useful in a log somewhere?

0 Likes
Highlighted
Absent Member.
Absent Member.

Generally what I will prefer below sequence when web client outage is reported,

 

  1. Web client
  2. SM fat client
  3. Network between SM web tier and User machine
  4. Network between SM and Web client server
  5. SM log files
  6. SM services
  7. SM , DB and web client system resources
  8. Network between SM and Database
  9. Rest depends on sm log file and type of mentioned issue details.
____________________________________
Assign Kudo, if found post useful and mark it accepted if solves the issue.
Highlighted
Absent Member.
Absent Member.

So, we did check web client

 

We forgot to check the fat client - sorry

 

What do you check regarding the networking?

 

In our case, since the web client server is the SM server, I think the networking between them is okay.

 

The sm log files show continued operation with no out of the ordinary warnings or errors.

 

The services were running - and recycling them did not help.

 

What specific "system resources" do you check? We checked the event logs and there wasn't anything extraordinary there.

 

From what we were able to find, there was no sign of the problem in any logs - yet no user was able to access the system through the web client until we rebooted the server.

 

 

 

0 Likes
Highlighted
Absent Member.
Absent Member.

What do you check regarding the networking?

Simply the network connectivity between user and Web(here SM server). It is quite possible that between remote user machine and web server there could be firewall or any other connectivity issue.

 

SM log files for any kind of licensing, high memory or number of limited connections etc.

 

System resources means CPU, memory or disk utilization etc. Bottleneck of any one of above can lead to service unavailability.

____________________________________
Assign Kudo, if found post useful and mark it accepted if solves the issue.
Highlighted
Absent Member.
Absent Member.

well, we RDP'd to the server from our desktop, so that connectivity was okay.

 

the log files showed nothing about high memory, connectivity, etc.

 

the cpu, etc. were all fine.

 

I appreciate your suggestions though!

0 Likes
Highlighted
Absent Member.. Absent Member..
Absent Member..

How many concurrent users are logging in?

How many login servlets are configured to start in the sm.cfg?

 

You may not be running enough login servlets for the total concurrent users. 

----------------------------------------------------
Kudos - what, where, how, and why
Want Good Answers? Ask Good Questions...
Highlighted
Absent Member.
Absent Member.

May be you can try login to server and check memory utilization of database server. Also check log file of database if it is allowing any user to make entry to it.

 

Servlet can also be the issue. Login to sm from application server itself giving localhost as address.

 

 

 

Regards,

Z

Highlighted
Cadet 2nd Class Cadet 2nd Class
Cadet 2nd Class

sometimes I wish it would disappear, sorry I know that doesn't help.

The opinions expressed above are the personal opinions of the authors, not of Micro Focus. By using this site, you accept the Terms of Use and Rules of Participation. Certain versions of content ("Material") accessible here may contain branding from Hewlett-Packard Company (now HP Inc.) and Hewlett Packard Enterprise Company. As of September 1, 2017, the Material is now offered by Micro Focus, a separately owned and operated company. Any reference to the HP and Hewlett Packard Enterprise/HPE marks is historical in nature, and the HP and Hewlett Packard Enterprise/HPE marks are the property of their respective owners.