High IOwiats on ELK Linux Instance

Hi All,

We have a Weblogic server and other applications/servers(HTTP server) that write logs to the Logs mount. In order to monitor the performance of the application, this mount is shared with the ELK stack.

After restarting ELK, we see high IOwiats on all Application Server instances where the /log mount is shared. Furthermore, ELK Linux instances have high IOwiat around the same time of restart.

ELK dockers are running on OEL 7.x ,16 core CPU and 31 GB memory

Elasticsearch is 3 node cluster .

logs mount is NFS mount

Is there a need to tune the ELK stack or how can we overcome this problem.

Thanks
Ak

Elasticsearch can be very I/O intensive and generally require fast storage in order to work well. Although it may be possible to run Elasticsearch on NFS it is generally not recommended as it can be very slow and limit performance. If you are indexing a lot of data it is recommended to use as fast storage as possible.

Thank you for your prompt response.