please increase CacheSize configuration parameter

[terrylin]

VMware Server
OS: RHEL 6.4
Database: Mysql
Zabbix Version: 2.4.4

I faced with a problem. Zabbix server wont start due to a memory error. logs details are below:

12602:20170625:172324.743 __mem_malloc: skipped 4 asked 29472 skip_min 448 skip_max 3032
12602:20170625:172324.744 file:dbconfig.c,line:446 zbx_mem_realloc(): out of memory (requested 29472 bytes)
12602:20170625:172324.744 file:dbconfig.c,line:446 zbx_mem_realloc(): please increase CacheSize configuration parameter
12595:20170625:172324.746 One child process died (PID:12602,exitcode/signal:1). Exiting ...

I already tried increasing CacheSize(4096M), TrendCacheSize, HistoryCacheSize (1024M or 2048M)and the server Memory(up to 8GB), and its still not working. the zabbix_server.conf setting about CacheSize is below:

VMwareCacheSize=2048M
CacheSize=6144M
HistoryCacheSize=1024M
TrendCacheSize=1024M
HistoryTextCacheSize=1024M
ValueCacheSize=4096M

The number of hosts is 211(146 running),
The number of items is 13800(13645 is running),
The number of triggers is 3625(3447 is ok),
When I enable 1 hosts, the service will not wok, search for many case's solution, but have no effect.

I would appreciate it that you give me some suggestions.
Many thanks.

[batchenr]

VMware Server
OS: RHEL 6.4
Database: Mysql
Zabbix Version: 2.4.4

I faced with a problem. Zabbix server wont start due to a memory error. logs details are below:

12602:20170625:172324.743 __mem_malloc: skipped 4 asked 29472 skip_min 448 skip_max 3032
12602:20170625:172324.744 file:dbconfig.c,line:446 zbx_mem_realloc(): out of memory (requested 29472 bytes)
12602:20170625:172324.744 file:dbconfig.c,line:446 zbx_mem_realloc(): please increase CacheSize configuration parameter
12595:20170625:172324.746 One child process died (PID:12602,exitcode/signal:1). Exiting ...

I already tried increasing CacheSize(4096M), TrendCacheSize, HistoryCacheSize (1024M or 2048M)and the server Memory(up to 8GB), and its still not working. the zabbix_server.conf setting about CacheSize is below:

VMwareCacheSize=2048M
CacheSize=6144M
HistoryCacheSize=1024M
TrendCacheSize=1024M
HistoryTextCacheSize=1024M
ValueCacheSize=4096M

The number of hosts is 211(146 running),
The number of items is 13800(13645 is running),
The number of triggers is 3625(3447 is ok),
When I enable 1 hosts, the service will not wok, search for many case's solution, but have no effect.

I would appreciate it that you give me some suggestions.
Many thanks.

i have the same number of hosts as you
this is my conf :
CacheSize=2G
HistoryIndexCacheSize=512M
ValueCacheSize=400M

all the rest is commented.
is your selinux is disabled\enforcing ?

[cros]

Continues ?

terrylin,

i have the same problem, Did you fix this problem ?

[terrylin]

[Solved]
Yes, I edited the wrong .conf, so it didnt work.

To find the correct .conf:
find / -name zabbix_server.conf

Edit the CacheSize from default to 32M.
Then check Monitoring -> Graphs -> Zabbix server -> Zabbix Cache usage graph

[cros]

Holes in the graphics

Terry lin
I know that you do not know me, but,
thank you friend, by the answer.

in my case, It was not only CacheSize I needed to increase.

I had to increase too.

- StartPollers=250
- StartPollersUnreachable=280

and

-max_connections in database mysql to 1000.

this solved to me too the holes in the all graphics.

To avoid the stop problem in the zabbix_server service after that time.
I created a task to start the zabbi_server service every day in the Linux task scheduler as a precaution

I hope this information will help others in a moment.

Sorry for my english and thank you terry.

[sph919]

Thanks Terry lin,
This also helped me

[cesarsj]

Our but what huge values yours, we have 862 hosts and 8652 items and our configurations are now:

StartPollers=5;
StartPollersUnreachable=2 (I think of increasing to 10);
CacheSize=16M (I think of increasing to 32M);

[charsit]

Thanks a lot to all for your comments, this helps me.