All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
VPS Outages
Hey folks,
Looking for some advice as I am having trouble figuring this out.
I have a VPS through umaxhosting.com that is running three WordPress sites with very low traffic <10 visitors a day. They are simple static sites with very little on them. I have it running CentOS and Centminmod.
I have setup 3 monitoring services on them, Pingdom, nodequery and Hetrixtools which all keep notifying me that the my site has been going offline. Ive noticed that my VPS will have high CPU usage and then go offline for a few mins although I don't see any of my processes causing the CPU spike. Example below.
My provider moved me to a new Node but it seems to still be happening. Not sure what to do or how to figure out if one of my processes is causing it, or an overloaded node.
Example email:
It seems one of your servers is consuming a lot of resources.
Last Update: 2018-11-01 02:21:33
Alert Trigger: 2018-11-01 02:22:02
Average: 4.40 2.97 1.50
System Load: 100%
Ram Usage: 22%
Disk Usage: 8%
php-fpm 2% 110.48 MB nginx (4)
mysqld 0% 29.29 MB mysql (1)
lfd - sleeping 0% 21.09 MB root (1)
nginx 0% 14.08 MB nginx (2)
crond 0% 3.96 MB root (3)
nginx 0% 2.83 MB root (1)
wget 0% 2.73 MB nodequery (1)
bash 0% 2.62 MB nodequery (2)
sh 0% 2.35 MB nodequery (2)
systemd 0% 2.09 MB root (1)
Comments
Migrate WP to Jekyll and host from Firebase
Make sure your VPS is not compromised
Is your site using PHP ? if not just disable php-fpm
As he mentioned Wordpress, I'm guessing he is using PHP.
OP, run "cat /proc/cpuinfo" and paste output. Also, make sure your WP has caching working i.e. keycdn cache enabler, wp super cache etc to off load some load from php-fpm.
This. Install Wordfence for WordPress temporarily and scan the files. Then uninstall.
Oops missed that, i just read static site
@prailone please run tcpdump -n and make sure you are not getting http flood or some kind of ddos attack even it can overload your VPS or even netstat -an can show if you are getting http flood
Move to other providers maybe?
If you don't see any CPU processes that bring the load up, it might signify that either
a) the load is caused by heavy IO and thus high storage latency, check with e.g.
iotopandiopingduring one of these spikes;b) it comes from another VPS/process on the host node itself, which momentarily starves your VPS of resources
I'm 99 percent sure this is OP fault. clearly he does not have enough memory and did not configure the vps properly so it is using swap heavily which will explain high cpu. Considering he did not even post his vps specs; not worth the time to help him.
Costly, if you need to move every day without solving the root cause.
Choosing a new host and moving works wonders.
Look at marriages.
Keep moving until you find one that sticks.
I mean there are tons of good providers here, why choose a CC brand instead
They are well-known for overloading CPU eh
swapping may be the reason for it to go offline.
Perhaps not your VPS, the node could have issues and swapping like crazy.
Might find better help on the centminmod forums.
Thanks everyone for the help and sorry for the late reply and not adding more details at the start. I have hosted a few simple WordPress sites before without any problems so this is driving me nuts especially since I haven't even launched the site yet. I'm using an under construction plugin as a placeholder for now.
Anyways here extra details to answer some of your questions:
Server details:
- 1GB OpenVZ
- Processor: Intel(R) Xeon(R) CPU E3-1240 v3 @ 3.40GHz
I have a few plugins installed to enable caching and secure the site:
- Auto optimize
- Cache Enabler
- CDN Enabler
- Sucuri Security (site is not compromised)
No traffic floods that I can see when doing a tcpdump or netstat
Ram usage is low:
https://imgur.com/Na1v1vg
Basically this.
(He needs to run iostat during a spike to see if there is some serious CPU steal.)
My idling VPS has similar problems too. I receive about 2 notifications about high system load per week even after I reinstalled the VPS and reset the monitor.
Sounds like it's an issue with the host node, to be honest.