- Server operating system version
- RockyLinux 8
- Plesk version and microupdate number
- 18.0.51 Update #1
The plesk update from this morning locked up two of my servers and a reboot was required to get them back online. The two servers in question were running RockyLinux 8 and Plesk 18.0.51 Update #1. I had two other servers running Centos 7 and the same update was applied. They were not effected. The monitoring log shows CPU usage sustained in high gear for the duration of time that it was locked up.
I am trying to trace the issue and prevent future problems.
One possible cause is that Port 53 (DNS) was closed to incoming traffic on the new servers. I thought since I wasn't running DNS I didn't need to have ports open for incoming traffic and up until now it didn't cause any issues. But after the update I could no longer use dig. Could Plesk have locked up trying to phone home after the update?
After the reboot I also had to rebuild my apache configuration file on one of the new servers. It only had one active site, which was a staging subdomain. I also couldn't connect to the staging site until I assigned it a Let's Encrypt cert. The default cert refused to connect. This wasn't the case before the update.
Anyone else have these issues? Any ideas on how I can trace this problem?
Thanks
I am trying to trace the issue and prevent future problems.
One possible cause is that Port 53 (DNS) was closed to incoming traffic on the new servers. I thought since I wasn't running DNS I didn't need to have ports open for incoming traffic and up until now it didn't cause any issues. But after the update I could no longer use dig. Could Plesk have locked up trying to phone home after the update?
After the reboot I also had to rebuild my apache configuration file on one of the new servers. It only had one active site, which was a staging subdomain. I also couldn't connect to the staging site until I assigned it a Let's Encrypt cert. The default cert refused to connect. This wasn't the case before the update.
Anyone else have these issues? Any ideas on how I can trace this problem?
Thanks