update tuned-2.11.0-12.el7_9.noarch cause Elasticsearch production crash

[love91k]

Steps to Reproduce:

1. update yum update -y, it will update tuned automatically to version tuned-2.11.0-12.el7_9.noarch
2. restarted the server of elasticsearch version 5.16.2 and found abnormal behaviour of elasticsearch.
3. yum downgrade tuned-2.11.0-12.el7_9.noarch, than restart elasticsearch sorted out the issue.

we found tuned recently updated our system as part of os activity and found tuned is automatically updated, because when we launch machine, we generally execute yum update -y to update all packages first and find out with latest tuned release iowait is drastically increased and cause full production cluster downtime. in checking in detail after multiple effort we have downgrade all package one by one which were upgraded in recent days and our system was fixed after downgrading to older version, there is some serious issue with tuned-2.11.0-12.el7_9.noarch

[yarda]

This is quite old TuneD release, the latest upstream release is tuned-2.19.0. Could you provide more info, what is causing the problem? I guess it could be something from the TuneD profile, not the TuneD itself. You could also try switching to the different TuneD profile or switching TuneD off. Although, we are ready to fix the problem upstream (if there is still one), we cannot get the fix to the RHEL-7. You will have to open downstream support ticket.