Spikes in system load on DO monitoring graphs

July 20, 2019 165 views
Apache Ubuntu 18.04

Ive noticed over the past 6 hours, a number of times, the DigitalOcean monitoring graphs are showing spikes of system load from 1 up to around 60, then a few minutes later dropping back to normal again.
During these load peaks the memory drops to 0 and some CPU spike also.

Ive noticed this across a number of my droplets around the same times today. Is this something I should look into? I dont think the site went down , not long enough to trigger an alert anyway

4 Answers

I am seeing the same thing. My best guess is that the monitoring software is losing its connection to the droplet or something. Nothing really explains why disk usage and memory would fall to zero during these events. Hopefully we'll get an answer from DO because this kinda freaks me out.

I am also seeing this too. Hearing it happen to others makes me think it's a monitoring/DO problem and not our servers. I've been monitoring via "htop" and my own logs, and have found nothing. Even during one the the actual spikes, I saw no change when SSH'd in. I imagine the the memory / cpu did actually hit 0, my programs would crash and I would see errors, but absolutely nothing.

I don't really mind it, but it would be nice to get an official response as to what it is. If it's a bug with monitoring, no problem, I'll just keep that in mind.

I'm the same...

https://ibb.co/Rz7qKTr

Assumed it was some sort of an attack. Would be nice if it was just an issue with the monitor.

I'm also having this issue and was concerned something was going wrong but it is reassuring I'm not the only person having this problem with the monitor. Hopefully it is just the monitor and nothing else?

  • I've not tracked any actual downtime. Nothing to indicate anything is wrong.

    I am still seeing the spikes however. No official word from DO, but with that being said, I haven't actually submitted a ticket.

Have another answer? Share your knowledge.