Office 365 probe is failing with inventory update message
Article ID: 136309
Updated On:
Products
Issue/Introduction
Office 365 probe is failing with the below error.
My-Profile-Name-Monitoring is not responding (reason: Connection and/or inventory update failure) {Read timed out}
When checking connectivity from the UIM Admin console interface for the Office365 probe, no error is seen.
Connectivity to https://manage.office.com:443 has been confirmed.
Error messages in the log seem to show that the inventory update succeeds but then immediately fails, e.g.:
Aug 20 09:33:26:241 [Data Collector - My-Profile-Name-Monitoring, office365] PERF: DONE: office365 inventory update for My-Profile-Name-Monitoring {Seconds=24.312}
Aug 20 09:33:26:241 [Data Collector - My-Profile-Name-Monitoring, office365] Inventory update failed. Sending resource alarm
Aug 20 09:33:26:242 [Data Collector - My-Profile-Name-Monitoring, office365] (1) error, Inventory update failed for My-Profile-Name-Monitoring: Read timed out
Environment
Release : 9.0.2
Component : UIM - OFFICE365
Cause
This is often a result of the memory settings for the probe being too low to handle the size of the monitored environment.
Resolution
Using Raw Configure or a text editor, edit the Offiice365 config file (office365.cfg).
In office365.cfg you will note the following which sets the initial and maximum memory values (the exact values may vary in your environment):
<startup>
options = -Xms32m -Xmx512m -Duser.language=en -Duser.country=US
</startup>
Make the change as follows (again, the numbers may need to be higher in your environment depending on what/how many nodes are being monitored.)
<startup>
options = -Xms1024m -Xmx1024m -Duser.language=en -Duser.country=US
</startup>
This will cause the probe to initially allocate the full amount of RAM instead of growing from 32mb, which will improve performance of garbage collection and memory usage. Increasing the memory in 512mb increments is suggested until a value is found which stabilizes the probe.
Feedback
