@mifiedle， Could you help test this bug and point out the steps? I have started 10 ocp-logtest pods with rate 600 on two nodes cluster. But I haven't watched the sessions on the scaled mux pods.
The easiest way to test is is with the ss utility, but unfortunately our ES pod image does not have it. I copy the utility there and run it. 1. From a linux system that has ss (from iproute package): oc scp /usr/sbin/ss -c elasticsearch <pod>:/tmp/ss 2. oc rsh into the ES pod 3. /tmp/ss -tnp | grep 9200 | awk {'print $5'} | cut -f4 -d":" | sort -u This shows all client IPs with connections to ES 4. Send log traffic for a while and repeat step 3. The list of connected clients should be different. 5. Check on all ES servers and each should have roughly an equal number of clients
Verified with logging:v3.10.45. The ES had roughly an equal connections.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2018:2660