Problems since WHM last update
Hi,
I have 2 servers running WHM 134.0.28. In both cases 2 major problems occurs since the last updates :
1) When I transfer accounts, for big accounts (several Gbs), the restore part get stuck at the Convert mail dir step. The conversion gets done, but it doesn't seem to go the the next step. I have to kill the conversion process manually in order to continue with the transfer. Before the update it was working fine, but was taking time, which was expected for big accounts.
2) Both servers get into high usage when I try to access my account list. Right now both servers are are high usage : 27+ for one, 9+ for the other. And in the process list nothing indicate a high usage.
The only way to make the server get back to normal usage is to reboot them. Not a realistic long term solution when you have clients counting in the server for their daily work.
I tried to fix all scripts using /scripts/upcp --force before coming to you.
Thanks for your help.
-
Hi,
For the transfer issue, do the transfer logs mention any warnings or errors before the process gets stuck? If the logs aren't helpful, you could try converting the mailbox format on the larger accounts before migrating. This can help to isolate any issues before beginning the transfer. We have documentation below on how-to manually convert mailbox formats:
Mailbox Format Conversion Scripts
It is unusual to see high resource utilization without any processes showing high usage. If you haven't already, I suggest using the sar command to check if there is a high %steal on the system, as this could cause high utilization without high usage processes. For reference:
What is CPU steal and when should I be concerned about it?
If the server also does not show signs of high %steal, I recommend opening a ticket so the issue can be looked into closer with server access.
0 -
Hi!
Thanks for your reply. I can provide with the log if you want to have precise data. As far as I can tell, there are errors but unrelated to the mailbox. A precision : I'm making transfers from Plesk servers to cPanel. As told before, there were issues, but nothing serious that was blocking the transfers.
As for the high usage, I will use the command you told me. I made some tests on my end and it seems that the error is linked to the second HDD that host my clients' accounts. This is odd because the configuration in my fstab has not changed for my first server.
0 -
Hi,Thank you for the additional details and context.
Regarding the transfer issue, performing an strace on the stuck processes can help to provide insight as to why the processes are stuck. However, it is often preferred to fully review the logs first since strace output can be excessive.
Additionally, since your source environment is Plesk, I wanted to mention that our migrations team does support assisted migrations from Plesk to cPanel. If you prefer a more guided approach or if issues persist, you may find this service helpful:cPanel Migration Services and Guides
With respect to the high load you’re observing, you finding that it correlates to the second HDD could be an important clue. I suggest checking for filesystem corruption as that can also be a cause of high load/IOWait. We have a guide on how-to check for disk corruption/filesystem errors below:
Identifying Disk Corruption and Filesystem Errors
0
Please sign in to leave a comment.
Comments
3 comments