New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Daily IO spike at VPS host?
Hi. I've recently moved to EDIS and their Swiss KVM location. Service looks great, but there appears to be a daily spike in IO, as can be seen under disk latency in in these Munin stats. I don't wanna take it to support just yet, so I thought I'd ask: Is this normal for VPS hosts? Do you reckon it's some low-priority background tasks running? I didn't see anything similar at my previous host, so I'm not sure what to think.
Comments
It's also a possibility that a handful of clients sharing your node may have been running IO tests around the same time. I know we get that occasionally whenever someone posts a new "test your VPS" thread.
Did you set the time according to your local time or to the time of the location of the server? I would guess it is because of cronjobs that are executed (stats get written, logrotation)
Daily spike at what hour exactly? The most probably culprit is people doing backups n_n
On http://178.209.51.9/munin/localdomain/localhost.localdomain/diskstats_latency/vda.html it looks like all of the spikes happen at approximately the same time every day without fail, so I'd guess that it's backups too.
Although measuring IO latency in "seconds" is pretty hellish.
@Damian: Yeah, I'm thinking backups also. But as you say -- SECONDS of IO latency? That's horrible. Maybe I should contact support about this.
@anincog: Yeah, i'd say so. Especially since you've had multiple spikes into the "seconds" range...
Seems that they're running some backups at the time of the IO peaks.