View Issue Details
|ID||Project||Category||View Status||Date Submitted||Last Update|
|0006268||CentOS-6||mysql||public||2013-02-23 23:37||2013-04-25 20:15|
|Priority||high||Severity||major||Reproducibility||have not tried|
|Target Version||Fixed in Version|
|Summary||0006268: MySQL 1023 connections limit|
a few weeks ago i changed the OS on my server from SuSE-Enterprise-Linux to CentOS 6.3.
Since then I`m having some issues regarding the MySQL server on it. The whole configuration from the old OS was just copied to the new server and adapted.
The problem is, that a maximum of 1023 connections ("Max_used_connections") to the server can be established while the limit in MySQL is set to 2800 connections.
What i`ve done so far (without improvement):
-Changing the "/etc/security/limits.conf"
-Setting some kernel parameters:
I`ve also found a Bugreport from RHEL4, in which a user was suffering the same problem, but I`m not able to find this report at the moment.
|Steps To Reproduce||As i don`t think you have any fat servers availible for just to reproduce this error, i may offer you to have a look via Teamviewer on this box if needed.|
Just drop me a mail to maximilian DOT herrmann AT jambit DOT com
mysql soft nofile 65536
mysql hard nofile 65536
mysql soft nproc 65536
mysql hard nproc 65536
mysql soft stack 65536
mysql hard stack 65536
ulimit -a as mysql:
[root@www56 ~]# su - mysql
-bash-4.1$ ulimit -a
core file size (blocks, -c) 0
data seg size (kbytes, -d) unlimited
scheduling priority (-e) 0
file size (blocks, -f) unlimited
pending signals (-i) 385852
max locked memory (kbytes, -l) 64
max memory size (kbytes, -m) unlimited
open files (-n) 65536
pipe size (512 bytes, -p) 8
POSIX message queues (bytes, -q) 819200
real-time priority (-r) 0
stack size (kbytes, -s) 65536
cpu time (seconds, -t) unlimited
max user processes (-u) 65536
virtual memory (kbytes, -v) unlimited
file locks (-x) unlimited
Kernel parameters optimized:
/proc/sys/net/netfilter/nf_conntrack_expect_max (now set to 4096)
/proc/sys/net/core/netdev_max_backlog (now set to 50000)
/proc/sys/net/ipv4/tcp_max_syn_backlog (now set to 30000)
/proc/sys/net/ipv4/tcp_max_tw_buckets (now set to 200000)
/proc/sys/kernel/shmmax (now set to 67645734912)
/proc/sys/kernel/shmall (now set to 16515072)
/proc/sys/net/core/somaxconn (now set to 65535)
|Tags||No tags attached.|
may anyone review this bug, please?
It`s a very unpleasant situation for me...
this is definitly a bug at CentOS, a collegue of me fixed this.
You have to write your limits down to /etc/security/limits.d /91-mysql.conf, because there`s the file "90-nproc.conf", which overrides all settings and doesn`t show up these limits on a ulimit.
It`s also possible to write these settings to a running process, for this you will need the PID and then do the following:
echo -n "Max processes=385852:385852" > /proc/<YOURPROCESSPID>/limits
Thanks for your REAL GREAT help!! Sorry, but i don`t understand this. It`s a serious problem for EVERYONE by using a bit more connections than the default settings, and nobody even answers with an question. I know that CentOS is an community product, but there are much more insignificant/nonrelevant "Bugs" (i mean some Layer-8-errors) which are getting much more RESPONSE than my question. I`m really nerved about this situation, and this is understated.