Since the exact switch of the month my CPU use jumped up by a big amount from around 5% before to around 27% after the switch of the month.
I double checked and it’s the storagenode process that’s using much more CPU. A restart of the container didn’t change anything. Any idea what might be going on?
I had that too but it only lasted like 30 minutes. Then no difference. At least I think it might be the same thing but then again, I am probably not in the same timezone and I didn’t check if that short spike actually was the storagenode.
That’s definitely not the same thing as for me it hasn’t dropped since. There are always tiny peaks around a whole hour, but it seems for me the base line has just increased.
I think I found the cause though, at that exact time a scheduled SMART extended test started. I think it’s probably slower IO causing an increase in CPU for my node. Ignore this post, it’s probably nothing, but I’ll report back when the test is done.
1 Like
Sounds likely.
I’ve seen memory increases as well on the docker containers for nodes, while storage has been more busy. This however, usually drops within an hour from normal conditions return.
Let us know how it evolves, could help us all build more in-depth knowledge and experience on what to watch out for 
As expected everything got back to normal when the SMART tests for each disk finished. Nothing to see here, apologies for the confusion.
2 Likes
Smart tests are hardware tests ran by Drive itself. Tests itself will not use any cpu.
No but the inscreased response time of the hardware might result in significantly higher iowait.
1 Like