High RAM Spikes
We are with Fasthosts as our dedicated server provider and we are running on Ubuntu 20. In the fast hosts panel, we are seeing our RAM (96GB) maxing out at 100% but the CPU usage stays low. At this time our websites stop loading. WHM and any monitoring software we have on there does not pick up these RAM spikes. We have contacted fast hosts support multiple times with nothing coming from that yet. We have to restart the server to bring the RAM back down again. This spikes between 20mins to 2hours after restarting. Any advice on this would be really appreciated.
-
Hey there! I would expect a *96G* RAM spike to be obvious if you are logged into the machine and checking either "top" "mysqladmin proc status" or the "apachectl fullstatus" commands - I would expect one of those to show something interesting related to that but you'd have to be watching the server in real-time and not relying on any log data in order to catch it in the act. 0 -
Hey there! I would expect a *96G* RAM spike to be obvious if you are logged into the machine and checking either "top" "mysqladmin proc status" or the "apachectl fullstatus" commands - I would expect one of those to show something interesting related to that but you'd have to be watching the server in real-time and not relying on any log data in order to catch it in the act.
Thanks for your reply. We have used the top command to monitor at the time we were having the problems but have not seen anything out of the ordinary. Here is an example of how it looked when we had the issue. I've also included the screenshots of load averages and IP connections which is what we see when it happens.0 -
Those screenshots don't indicate any issues with the machine. In fact, the "top" screenshot shows there is 76G of RAM available on the system during that time. In your original post, where did the image of the RAM usage come from? 0 -
The green graph is from monitoring software we installed on WHM which is showing almost no usage while the blue graph is from our fasthosts control panel. 0 -
I'm not sure if I believe the blue graph if it contradicts what you're seeing directly on the server over SSH. 0 -
Yes I am inclined to believe that this graph is either a red herring to the real server issue or there is some kind of hardware fault. I have just got a reply from a cPanel ticket explaining that they have found the server is reaching the "maxRequestWorkers". This does fall in line with the problems we are facing. I am investigating this now and I'll reply if I find this to be the problem. 0 -
If you already have a ticket opened with us that'll be much better than me guessing over Forums :D although I will say the initial command I mentioned of "apachectl fullstatus" would have shown a full scoreboard and revealed the issue. 0 -
Ok thank you so much for your time. 0 -
Sure thing! 0
Please sign in to leave a comment.
Comments
9 comments