temporary system logs eating up disk space

[gargi_talukdar]

We have recently installed APIGEE Edge private cloud version 4.18.01 and seeing an issue where huge temporary logs are getting created under edge-message-processor (/apigee/var/log/edge-message-processor/logs) which is eating up the disk space.

Example:

-rw-rw-r--. 1 apigee apigee 105640677376 Feb 16 09:12 system.log6044788046682057.tmp

This is happening even though the instance is not in use currently.

Also I am seeing the error after the disk space is getting all used up.

Error while accessing datastore;Please retry later.

[API-Evangelist]

Why would it create .tmp file? Usually it would .gz the files and rotate. We don't have such symptoms.

Can you stop MP & remove the .tmp file and do a restart?

==

Error while accessing datastore;Please retry later

==.

Is your ZK/CS are up and running? Can you run a telnet from MP to ZK/CS host/port & verify?

-Vinay

[#3]

I have the same problem with version 4.18.01. The system logs were full of "java.io.IOException: Too many open files". My solution:

apigee-all stop
cd /opt/apigee/var/log/edge-message-processor/logs
rm -f system*.tmp

check this:

https://community.apigee.com/articles/1422/apigee-edge-opdk-file-descriptors-recommendations.html

I have added the following line to the end of /etc/security/limits.conf :

*     hard    core     0

Reboot.

I hope this helps.