Can I reduce beam.smp thread numbers due to load average high?


#1

There are only 4 cores, so load average load average: 6.17, 3.65, 2.81 is high, and I want to know about the detail that beam.smp is doing(I Know: These processes are responsible for monitoring and managing all other underlying server processes such as ongoing XDCR replications, cluster operations, and views.)?

Can I change the beam.smp thread numbers?

OS : CentOS release 6.7 (Final)
Version: couchbase-server-community-4.0.0-4051.x86_64
Cluster: 3 x c3.xlarge (14 ECU, 4 vCPU, 2.8 GHz, 7.5 GiB Memoey)

top -H -i

top - 11:15:13 up 11:00,  0 users,  load average: 6.17, 3.65, 2.81
Tasks: 368 total,   7 running, 361 sleeping,   0 stopped,   0 zombie
Cpu(s): 28.9%us,  6.0%sy,  0.0%ni, 43.1%id, 22.0%wa,  0.0%hi,  0.0%si,  0.0%st
Mem:   7017368k total,  4138720k used,  2878648k free,   177740k buffers
Swap:        0k total,        0k used,        0k free,   585140k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
 1710 couchbas  20   0 3358m 2.0g 3360  R 72.2 29.3 134:59.01 beam.smp
 1712 couchbas  20   0 3358m 2.0g 3360  R 48.5 29.3 134:12.02 beam.smp
 1711 couchbas  20   0 3358m 2.0g 3360  R 33.6 29.3 134:41.51 beam.smp
 1713 couchbas  20   0 3358m 2.0g 3360  R 19.8 29.3 133:41.65 beam.smp
 2153 couchbas  20   0 1280m 185m 9820 R  1.0  2.7   6:31.61 beam.smp
26410 root      20   0 15300 1460  924 R  1.0  0.0   0:51.88 top
  421 root      20   0     0    0    0  D  0.0  0.0   0:33.59 jbd2/sda1-8
 1694 couchbas  20   0 3358m 2.0g 3360  D  0.0 29.3   0:09.57 beam.smp
 1697 couchbas  20   0 3358m 2.0g 3360  D  0.0 29.3   0:10.27 beam.smp
 2152 couchbas  20   0 1280m 185m 9820 R  0.0  2.7   6:57.15 beam.smp

#2

Did you figure out what the issue was?