WLS Admin Console Became Extremely Slow (Doc ID 1604558.1)

Last updated on MARCH 06, 2024

Applies to:

Symptoms

The WLS admin console became extremely slow during live operation, which halted necessary application EAR file deployment activities. When it happened, the page loads take up to 10 minutes to load (apparently at random times but most of the time), especially the ones involving checking the health status of other servers.

Strangely, the number of active nodes was not showing correctly (4 instead of 9, sometimes 0) despite that the WS service URLs were accessible on all nodes (and external clients were connecting to the managed servers fine).

Some admin pages load quickly sometimes, especially the ones not involving communication with other server nodes.

Several things were tried to resolve the issue:

1. Restart the admin server node; done; no results.
2. Restart the admin machine; done; no results.
3. Check the logs after admin start; done; no results (normal logs).
4. Restart all node manager services; done; no results.
5. Increase admin node heap memory; done from 512mb to 1024mb; no results.
6. Clear admin node tmp and cache folders; done; no results.
7. Ping and telnet from admin machine to other machines on node manager ports and server ports using listen IPs configured; done; normal communication.
8. Restarting all server machines, physically did resolve the issue; the admin became fast again, but the root cause was not found.

Cause

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.