Replies: 1
Hey, guys, I am a bit stumped here and hoping to get some advice from the community.
I have an existing WordPress server running on an LAMP Turnkey VM deployed from an OVA onto a home lab VMware ESXi server. Its been running a personal blog for 3 years without any problems. I’ve seen the ‘Error Establishing a Database Connection’ a couple times in the past which was resolved by restarting the server, I didn’t think much of the problem as the blog is a hobby and not a big deal if its down for a day or two before I realise there is an issue.
So 3 days ago the same error appeared, unfortunately, this time a restart did not fix the problem. I spent a little bit of time troubleshooting which I go into detail below on without any luck so I ended up restoring the entire VM from a backup. Luckily my content doesn’t change much as the closest restore point was 25 days ago. Weird thing is, the newly restored WordPress VM experienced the exact same error. Weird because the restore point was 25 days old and it only mucked up a day ago, this leads me to believe the problem may be an external factor and not the VM itself.
Here are my notes on troubleshooting.
1) Does the problem occur for /wp-admin/ , Yes it does
2) Performed repair and optimise on database, done
3) checked DB username / password, yep all good
4) check MySQL is running, the service has stopped and needs to be started, I start the server it may run for 10 minutes or an hour before it stops again.
5) check resources, CPU and RAM usage is maxed out reported by VMware, Debian OS is reporting 0.97 CPU but low RAM usage. The disk is SSD and low to non latency reported.
6) ran up another turnkey LAMP and imported DB and wordpress, same issue
7)ran up another brand new turnkey LAMP and started on a new WordPress from scratch, it was ready for content to be created and I activated the port forwarding for port 80 to internal IP of the LAMP wordpress server, CPU usage jumped to 100% immediately, note this is with a brand new DB/WordPress install on a brand new Debian OS LAMP server… no content was imported. If I remove the port forwarding then the CPU drops to 1%. These happens to the old wordpress server as well.
So I am kind of stumped, I checked if it was DDOS and I could only see half a dozen connections. It’s clear as day whenever I enable the port forwarding for port 80 to the internal IP of the TurnKey LAMP server running wordpress, CPU usage jumps through the roof.
-
This topic was modified 24 minutes ago by
rhyshammond.
