Doshboard timeout while refreshing

multix

Hi there. I have a problem with netxms server 2.1 Sometimes (several times in 1 hour) Dashboards are getting timeout error while refreshing. But connection is not breaking down and a litle later, dashboard can refresh everything with no error.

I am using dedicated physical server pc for netxms server. Specs :
2 * Sockets with 12 Cores (total 48 cores with hyper threading enabled)
4* SSD Drive with raid,
128 GB Ram,
Centos 7,
Postgresql 9.6.3

Top Command result :
Tasks: 507 total, 3 running, 503 sleeping, 0 stopped, 1 zombie
%Cpu(s): 2.6 us, 3.3 sy, 0.0 ni, 92.4 id, 1.5 wa, 0.0 hi, 0.1 si, 0.0 st
KiB Mem : 13189904+total, 2368860 free, 6030832 used, 12349934+buff/cache
KiB Swap: 4194300 total, 4194300 free, 0 used. 11046916+avail Mem

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
15841 root 20 0 26.819g 2.455g 5564 S 238.9 2.0 2755:30 netxmsd
4443 postgres 20 0 10.492g 905084 889572 R 31.6 0.7 0:58.43 postmaster
7128 postgres 20 0 10.506g 61756 34000 R 9.6 0.0 0:00.29 postmaster

As you can see, OS is not swapping. And normally, everything is very vell. I can see line chart that has 8 lines and 60 secs data collection time with 30 day in 3-4 seconds.

Other Information about netxms server and pc is :

Database size : 116-120 GB

netxmsd: sh dbstats
SQL query counters:
Total .......... 464511464
SELECT ......... 108457658
Non-SELECT ..... 356062135
Long running ... 0
Failed ......... 1
Background writer requests:
DCI data ....... 122998830
DCI raw data ... 123126498
Others ......... 32500

netxmsd: sh stats
Total number of objects: 31872
Number of monitored nodes: 1601
Number of collectable DCIs: 33637

netxmsd: sh qu
Data collector : 0
DCI cache loader : 0
Database writer : 0
Database writer (IData) : 4
Database writer (raw DCI values) : 1
Event processor : 0
Node poller : 0
Syslog processing : 0
Syslog writer : 0

netxmsd: sh watch
Thread Interval Status
----------------------------------------------------------------------------
Item Poller 10 Running
Syncer Thread 30 Sleeping
Ad hoc scheduler 5 Sleeping
Recurrent scheduler 5 Sleeping
Poll Manager 5 Sleeping

I will send an other post to add more screen shots.

By the way, this timeout problem was not happening in netxms 2.0.8. It started when I upgraded to 2.1 from 2.0.8

Can you help me, please. Thanks.