Answer the question
In order to leave comments, you need to log in
linode and swap
Recently migrated one project from AWS to linode.com. In the course of using their servers, I had a feeling, backed up by facts, that if the system starts using swap (even if by some small percentages), I / O immediately jumps beyond the reasonable, everything starts to go dumb. For myself, I solved this by completely disabling swap (well, adding memory), but I would like to know if there are other solutions, or is there really a disk subsystem that is such a bottleneck
PS I understand that swap is bad, but so on server performance it hasn't affected me yet.
PPS I use the standard ubuntu from their defaults.
Answer the question
In order to leave comments, you need to log in
I once met a problem with new disks with 4k formatting, if the alignment is wrong, then any good load on the disk outputs up to 100 percent in io.
As a result changed screws on the server.
I met this problem with keyweb.
Lay out the screw model, you can see it in parted / dev / sda
Didn't find what you were looking for?
Ask your questionAsk a Question
731 491 924 answers to any question