Iowait

[wuurp]

So what exactly does 5028% iowait mean?

[Andre]

I've just checked this with SWsoft and they told me this could possibly be a bug in Virtuozzo. If you could open a support ticket with your VPS details we can investigate this.

[wuurp]

It's ticket LPL-23471 that I opened a couple days ago. Still waiting for an update from Milovan.

The top screencap was while one person was connected to the vps and making an http download. It spiked like this several times while I was watching, mostly to just 2 or 3000, and possibly the same cpu each time. It would also switch from say 5000% iowait to 5000% idle and back.

[Andre]

Hi wuurp,

What you say there is not correct. As you can see in that screenshot there are lots of httpd processes running, so there are much more people connected to your VPS.

As said, please discuss this in the trouble ticket, on here there's nothing we can change about it.

[wuurp]

Well, no, it was just one person, me. I specifically waited until no other connections were active. Netstat showed my http and ssh connections. Nothing else. I switched top to showing threads, so you're seeing the various httpd threads there.

[helmishariff]

What is really happened? My vps iowait 100% even cpu is idle. A few hour after my vps upgraded to new virtuozzo ver.

ticket id: ASS-11776

top
---
04:44:11 up 9:19, 2 users, load average: 5.75, 4.18, 2.56
80 processes: 78 sleeping, 1 running, 1 zombie, 0 stopped
CPU0 states: 0.0% user 0.0% system 0.0% nice 100.0% iowait 0.0% idle
CPU1 states: 0.0% user 0.3% system 0.0% nice 99.2% iowait 0.0% idle
CPU2 states: 0.0% user 0.0% system 0.0% nice 100.0% iowait 0.0% idle
CPU3 states: 0.0% user 0.0% system 0.0% nice 100.0% iowait 0.0% idle
Mem: 8300388k av, 148552k used, 8151836k free, 0k shrd, 0k buff
7062204k active, 593884k inactive
Swap: 8193140k av, 16k used, 8193124k free 0k cached

PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME CPU COMMAND
1 root 16 0 1316 488 1276 S 0.0 0.0 0:00 3 init
2013 root 15 0 1368 544 1320 D 0.0 0.0 0:00 2 syslogd
2024 nobody 16 0 4268 2196 3680 S 0.0 0.0 0:00 2 proftpd
3075 root 15 0 3460 1488 3348 S 0.0 0.0 0:00 0 sshd
3486 root 15 0 1368 592 1316 S 0.0 0.0 0:00 2 crond
3513 root 18 0 2048 1048 1936 S 0.0 0.0 0:00 1 mysqld_safe
3553 mysql 15 0 34200 14M 3760 S 0.0 0.1 0:00 3 mysqld
3797 xfs 16 0 6856 5612 1972 S 0.0 0.0 0:00 2 xfs
4009 root 15 0 1332 488 1284 S 0.0 0.0 0:00 3 portsentry
20093 root 16 0 2236 1372 1936 S 0.0 0.0 0:00 2 bash
20430 root 15 0 9360 8144 2628 S 0.0 0.0 0:00 2 chkservd
23978 named 15 0 21296 4080 4676 S 0.0 0.0 0:00 2 named
28135 root 34 19 8396 6912 3308 S N 0.0 0.0 0:00 1 cpanellogd
28177 mailnull 16 0 5376 3952 2984 S 0.0 0.0 0:00 0 eximstats
28233 nobody 18 0 3600 2360 2440 S 0.0 0.0 0:00 1 entropychat
9956 root 16 0 2644 1116 2344 S 0.0 0.0 0:00 1 postsuexecins
9969 root 16 0 6932 4632 4732 S 0.0 0.0 0:00 0 httpd
9977 root 15 0 4740 3316 2892 S 0.0 0.0 0:00 0 leechprotect
20098 postgres 16 0 7476 1928 6904 S 0.0 0.0 0:00 0 postmaster
20100 postgres 16 0 8452 1908 6904 S 0.0 0.0 0:00 3 postmaster
20101 postgres 16 0 7508 1944 6904 S 0.0 0.0 0:00 3 postmaster
23923 root 16 0 22848 20M 2000 S 0.0 0.2 0:00 1 clamd
23933 mailnull 16 0 6648 1812 6008 S 0.0 0.0 0:00 0 exim
23937 mailnull 18 0 6604 1784 6008 S 0.0 0.0 0:00 3 exim
23964 root 16 0 2768 1432 2344 S 0.0 0.0 0:00 0 antirelayd
24181 root 15 0 9912 8624 7628 S 0.0 0.1 0:00 1 cppop
24557 root 16 0 26124 21M 5536 S 0.0 0.2 0:00 0 spamd
24567 root 16 0 28780 24M 5536 S 0.0 0.3 0:00 1 spamd
9541 cpanel 18 0 3776 1620 3376 S 0.0 0.0 0:00 1 stunnel-4.04l
9547 root 15 0 11648 10M 9124 S 0.0 0.1 0:00 0 cpsrvd
14154 root 16 0 27448 23M 5536 S 0.0 0.2 0:00 2 spamd
17410 root 16 0 1968 836 1684 S 0.0 0.0 0:00 1 xinetd
24088 root 16 0 2216 1316 1936 S 0.0 0.0 0:00 0 bash

[RaZor Edge]

Seems like you have the same problem that I have since two month now:

http://www.webhostingtalk.com/showthread.php?t=502547

Even with no connections to my VPS (apache and mysql stopped), 3 or 4 time a day, IOwait will get stuck at 100% for about 20/40 minutes (this situation is not related to my website activity). Any activity who need to write to the disk will cause this high IOwait.

When this problem is not happening, i've got server load average of 0.20/0.40 and my VPS is really fast.

Since thursday, Milovan is looking into this (Ticket FCE-48164).

[wuurp]

You think you've got problems

avg-cpu: %user %nice %system %iowait %steal %idle
0.00 0.00 0.00 107454770.73 0.00 102.75

107454770.73%

Ok, just kidding, that's obviously a bug.

For me, the iowait doesn't stay high, but it bounces around a lot, and often will hit 100% on all 4 for brief periods. I'm trying to monitor this over a few days and see if I notice any real pattern.

[TomH]

I have these problems too, I think those IO issues are related to my slow querries which are still not fixed

[Andre]

If anyone is having unexpected IOwait then feel free to open a support ticket then we will investigate this for you and make suggestions based on our observations.

[helmishariff]

Yes, better you guys open a support ticket. My problems have been solved. It is cause by conflict configuration file. That is told by the support guy. For now my vps is stable. Anyway, my vps performance much better now after virtuozzo has been upgraded yesterday. First benchmark 186 unixbench point. Second 12 hours after problems have been solved, 163

[Andre]

That's good to hear