Skip to main content

Server hangs during nightly maintenance

Comments

9 comments

  • cPRex Jurassic Moderator
    Hey there! I have a few thoughts on this. The first thought, is that the excessive warnings from LFD can be ignored. Those are just telling you a process ran longer than their default setting of 1800 seconds, which is completely possible with those main PHP processes. Those likely wouldn't not be related to your issue. That 4th block of text you posted shows there there was an OOM event for the system overall, so something is causing the machine to run out of memory, and also to experience a high load condition. I would recommend checking the sar logs on the system to see if that helps you identify any specific issues on the machine: If your server consistently has this issue overnight at the same time, it might be simplest to just watch the server in real-time so you can see what happens. However, if the system gets to the point where even SSH does not work, the issues indicate a deeper problem than the cPanel software.
    0
  • ddaddy
    Thanks for the reply. All `sar` gives me, regardless which file I choose, is: Linux 4.18.0-425.19.2.el8_7.x86_64 (myserver..com) 04/24/2023 _x86_64_ (2 CPU) 06:26:19 LINUX RESTART (2 CPU)
    As it's an AWS EC2 instance, i'd have thought any HDD issues would be flagged to them internally, but it's possible I guess. When I say SSH doesn't work, I think it's just a case of OOM like you say stopping it connecting/working properly.
    0
  • cPRex Jurassic Moderator
    That's odd - I would expect the sar logs to have much more data there. We'd be happy to take a look in a ticket if you'd like to submit one, but I'm not sure there's going to be much for us to check if those logs don't exist. It might be best to reach out to the hosting provider to see if they can check the system, or if they have a way to do more advanced monitoring at that time. I still think the best action to get the most useful details would be to watch the server load in real-time during the time the issue happens.
    0
  • ddaddy
    Thanks. I'm going to gather a bit more info by changing the times of the cron jobs. Currently update is running just before midnight, and backup at 2am. So it could be either of these causing the issue. I've changed them 2am and 9am and will come back if I get the same thing happen.
    0
  • sparek-3
    What specific distribution are using? What kind of virtualization is AWS using these days? [font="courier new">dmidecode -s system-product-name
    0
  • SimpleSonic
    I"d be interested to see the Apache logs as well as any netstat output around the time it starts going south to rule out any possible layer 7 type attack.
    0
  • ddaddy
    What specific distribution are using?

    AlmaLinux release 8.7 (Stone Smilodon) [QUOTE] What kind of virtualization is AWS using these days? dmidecode -s system-product-name
    This just returns the instance type t3.small
    I"d be interested to see the Apache logs as well as any netstat output around the time it starts going south to rule out any possible layer 7 type attack.

    The logs have been deleted due to The "Delete each domain"s access logs after statistics are gathered" option is enabled
    but if it happens again i'll be sure to grab them.
    0
  • ddaddy
    It's been a couple of weeks and this has started again. It now happens just after 9am which to me confirms it has something to do with the upcp Cpanel update as I changed it to run at 9am. The last 2 days at approx 9:05 I am notified that my websites are down. The CPU's max out causing everything to be unresponsive and it stays like that until I reboot the server.
    0
  • cPRex Jurassic Moderator
    Can you submit a ticket to our team so we can take a look?
    0

Please sign in to leave a comment.