Bug#679723: mysqld: started using loads of CPU when leapsecond passed

June 30th, 2012 - 11:00 pm ET by Julian Gilbey | Report spam
Package: mysql-server-core-5.5
Version: 5.5.24+dfsg-4

My mysqld recently started using loads of CPU, on two different
computers. There was no obvious reason, and restarting the daemon did
not appear to help. Finding a thread about mythtv suggested a
possible reason: the leap second which occurred at 23:59:60 GMT on 30
June 2012 (about 3 hours ago now). Rebooting the computer seems to
have sorted the problem.

Obviously, this is going to be hard to reproduce and test - either it
will have happened to lots of people or not, and it probably won't be
reproducible until the next leap second occurs :-(

I'm running kernel 3.2.0-2-amd64 (linux-image 3.2.20-1) and running
ntpd.

Julian



To UNSUBSCRIBE, email to debian-bugs-dist-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
email Follow the discussionReplies 3 repliesReplies Make a reply

Replies

#1 Nicholas Bamber
June 30th, 2012 - 11:40 pm ET | Report spam
Actually on my Linux system it the mysqld process has the hghest CPU:

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND

2560 mysql 20 0 408m 2372 692 S 53 0.1 155:39.60 mysqld

On kreebsd-i386 its low
74852 mysql 128 0 223m 32m 0 S 0.0 1.6 3:30.71 mysqld



I tried an strace

:/tmp# strace -p 2560
Process 2560 attached - interrupt to quit
restart_syscall(<... resuming interrupted call ...>

I'm thinking what else I can do to investigate.

On 01/07/12 03:55, Julian Gilbey wrote:
Package: mysql-server-core-5.5
Version: 5.5.24+dfsg-4

My mysqld recently started using loads of CPU, on two different
computers. There was no obvious reason, and restarting the daemon did
not appear to help. Finding a thread about mythtv suggested a
possible reason: the leap second which occurred at 23:59:60 GMT on 30
June 2012 (about 3 hours ago now). Rebooting the computer seems to
have sorted the problem.

Obviously, this is going to be hard to reproduce and test - either it
will have happened to lots of people or not, and it probably won't be
reproducible until the next leap second occurs :-(

I'm running kernel 3.2.0-2-amd64 (linux-image 3.2.20-1) and running
ntpd.

Julian



_______________________________________________
pkg-mysql-maint mailing list

http://lists.alioth.debian.org/cgi-...ysql-maint







To UNSUBSCRIBE, email to
with a subject of "unsubscribe". Trouble? Contact

Similar topics