Hello CA-7 / iDash Users:
I have a question about the inactivity timeout limit in iDash.
Our intent is to use the iDash monitor in the command center where it can be monitored and viewed by the Data Center team.
I understand the inactivity timeout can be increased in the idash.conf file under the idash.ui.session.inactivity.timeout= xx.
However, by increasing this amount , we encountered unpredicted results as identified by CA. We encountered multiple false positive missed slas. We have since reverted back to the default time of 30 mins.
We also, considered the refresh rate, but as I understand it in CA7 iDash, this refers to the job SLA refresh.
Does anyone have ideas or recommendation on how they monitor iDash with extended limit or refresh? (Release 12.0.04)
I would appreciate any feedback or advise.
Incorrect SLA processing isn't a side-effect of the inactivity timeout value. Can you please elaborate on "as identified by CA"? Was this suggested to you by us in response to a support ticket that you may have opened up for this? We have customers monitoring the dashboard page by specifying a high value for the UI timeout property without an impact to the SLA processing.
Thanks for the reply. The case I had opened with CA is 00749479. The actual response from the resolution was " you could end up crippling iDash to where it does not function properly"
We had iDash not reporting any false positives for quite sometime. After increasing the time limit, we started having the false positives again.
The comment made on that case should have shared that if you have a lot of active iDash users (more than 50) that are actively on Web UI pages that refresh periodically then that would be an overhead for the iDash server.
If you increased the timeout limit and noticed false positives then please open a ticket with a link to this post so we understand the background. We will then look at your iDash environment to understand the cause of the behavior. The hardware configuration will also be considered.
Is 24 hrs or less reasonable?
In past cases I understand that there were logs required. Just to be prepared in case this happens again, which type of logs and timeframe (mins/hrs before/after the sla is missed) will be required?
1440 minutes as the timeout value is reasonable.
To understand the reason behind the SLAs behaving in an odd manner, we would require all the files from IDASH_HOME/log/sla, alarm and archive(sub-directories under sla) directories that begin with the name of the SLAs in question. We would also require all the IDASH_HOME/log/info* and error* files. For the SLAs in question, the screenshots of the SLA List page sorted with these SLAs visible and the page where the SLA details (for the top two SLAs) are displayed would be helpful.