Results 1 to 4 of 4
Enjoy an ad free experience by logging in. Not a member yet? Register.
- Join Date
- Aug 2013
Linux server crashes at random intervals
Me and my friends have a linux CentOS server running. We use this server to run our minecraft server on. We are now using this server for about a week. But at random intervals (2hours-2days) the server suddenly crashes. I have tested to run the minecraft server at my personal computer, but that did not give any results. Note that the actual hardware server is down, with that taking our minecraft server also down. I have to reset the server by using a 'automated hardware restart' in the host company their website (Hetzner).
The messages file says the following:
Anyone any idea what the problem could be?
Thank you for reading my post, plz help!
- Join Date
- Sep 2007
- Silver Spring, MD
Items to review
1st - I think the first thing you want to do is to capture the logs.
2nd - Run the command kdump - Collecting and analyzing Linux kernel crashes - Kdump
3rd - Next, evaluate all of the items on the system from a hardware standpoint (from memory, hard-drive, cpu, network cards, make sure that everything has been seated
4th - Check the crontab or cron jobs to view if there is something being loaded there.
5th - Run top command to see what is running on the machine
6th - Run lsof -iTCP -P -s - Identify network processes
7th - Disconnect the system from the network and verify if there is an outside connection affecting the system
8th - Identify any errand processes (ps -ef)
9th - Stop any processes that are not essential to the system- chkconfig --level 345 <proc> off, once you do that, you should be able to normalize the system.
10th - If it has been identified as being a virus or malware, then run download avg - http://www.avg.com/us-en/avg-linux-email-server-edition
1. Keep the temperature in the premise where the server is within the specified by the OEM temperature range.
2. Try to improve the ventilation of the server ... & of the working premise.
3. You may try also to replace the air-cooling of the CPU with water cooling.
4. Reduce the clock of the processor (if this option is available)
5. Make a survey of the motherboard - with a magnifying glass. Watch for cracks on the bus rails, carbonification of chips (as a result of overheating), short-circuits from contamination with dirt, etc.
6. Clean up the motherboard.
For further details of how to do this see: How to Clean a Motherboard | Cleaning Guides
- Join Date
- Oct 2012
I am thinking that the problem lies in the MTRR cleaner. There are numerous lines in the log you posted that shows the kernel trying to find the most optimal configuration (for example: gran_size: 2G chunk_size: 2G num_reg: 4 lose cover RAM: 1950M)
You could try to manually tune MTRR by adding the following to the boot parameters: