First introduce the next kernel.sem this parameter
CAT/PROC/SYS/KERNEL/SEM250 32000 32 128 or IPCS-LS------Semaphore Limits--------max number of arrays = 128max semaphores per array = 250max semaphores system wide = 32000max Ops per semop call = 32semaphore max value = 32767 signal on my machine Queue Arrays Ipcs-a | wc-l144144>128 Why super, this I do not understand, but my openmanage plugin has been timed out, and can not start, can not stop, real TM Check this problem took me a long time
First, the meaning of the 4 parameters of the next kernel.sem is introduced.
cat/proc/sys/kernel/sem250 32000 128250 semmsl max semaphores per array signal set accommodates maximum number of signals 32000 Semmns Max semaphores System wide maximum number of all signals (SEMOPM Max OPS per semop call) maximum number of signals in a single signal set 128 Semmni max number of arrays signal set maximum
This is on the Internet, but it feels not right, according to our thinking
First of all
Max semaphores per array
Max number of arrays
These 2 parameters are definitely not enough, let's start with a little higher
echo "641280" >/proc/sys/kernel/sem
It's done.
We use
ipcs-a | grep Nagios | Wc-l90 view current Nagios user signal count
And then clean it all off.
ipcs-a | grep Nagios | awk ' {print $} ' | Xargs-n1 ipcrm-s Explanation of the 2 commands on IPCS and Ipcrm I'm not going to say any more online.
ipcs-a | grep Nagios | Wc-l0
After all, restart the application such as
/opt/dell/srvadmin/sbin/srvadmin-services.sh stop
Kernel.sem semaphore parameter tuning, and IPCS semaphore queue query