Re: High system times on Scaleway Cloud server

[Alexander Nasonov <alnsn%yandex.ru@localhost>]

Alexander Nasonov wrote:
> (taken ~1 hour later)
> # pstat -T         
> 624/98304 files
>  366840 vnodes
> 0/9585162 blocks swap space

I don't know if it's relevant but a number of vnodes slowly grows:

(taken ~45 minutes later)
# pstat -T 
624/98304 files
 376602 vnodes
 0/9585162 blocks swap space

I watched /usr/bin/top for a while but I didn't spot any vdrain
wchan from my previoius email:

>   PID USERNAME PRI NICE   SIZE   RES STATE      TIME   WCPU    CPU COMMAND
>     0 root     125    0     0K  358M vdrain/0  66:39  3.66%  3.66% [system]

I see many pgdaem:

  PID USERNAME PRI NICE   SIZE   RES STATE      TIME   WCPU    CPU COMMAND
    0 root     126    0     0K  365M pgdaem/0  72:09  7.67%  7.67% [system]

and occasional syncer:

  PID USERNAME PRI NICE   SIZE   RES STATE      TIME   WCPU    CPU COMMAND
    0 root     124    0     0K  369M syncer/6  72:32  3.52%  3.52% [system]


The above with more details:

load averages:  14.4,  13.2,  11.8;               up 1+13:36:48                                                           15:19:06
130 processes: 9 runnable, 112 sleeping, 9 on CPU
CPU0 states:  7.5% user,  0.0% nice, 86.6% system,  6.0% interrupt,  0.0% idle
CPU1 states: 18.4% user,  0.0% nice, 80.6% system,  0.0% interrupt,  1.0% idle
CPU2 states: 12.4% user,  0.0% nice, 87.1% system,  0.0% interrupt,  0.5% idle
CPU3 states:  5.5% user,  0.0% nice, 94.0% system,  0.0% interrupt,  0.5% idle
CPU4 states:  6.0% user,  0.0% nice, 92.5% system,  0.0% interrupt,  1.5% idle
CPU5 states: 19.4% user,  0.0% nice, 79.6% system,  0.0% interrupt,  1.0% idle
CPU6 states: 10.0% user,  0.0% nice, 89.6% system,  0.0% interrupt,  0.5% idle
CPU7 states:  7.5% user,  0.0% nice, 91.0% system,  0.5% interrupt,  1.0% idle
Memory: 3533M Act, 2041M Inact, 7572K Wired, 34M Exec, 5103M File, 620M Free
Swap: 9361M Total, 9361M Free
Seconds to delay: 
  PID USERNAME PRI NICE   SIZE   RES STATE      TIME   WCPU    CPU COMMAND
    0 root     126    0     0K  365M pgdaem/0  72:09  7.67%  7.67% [system]
10639 alnsn     77    0    10M 2012K wait/1     0:00  8.05%  1.12% sh
 9484 torpubli  85    0    64M   53M kqueue/0   4:09  1.03%  1.03% tor



load averages:  8.86,  10.0,  10.8;               up 1+13:44:44                                                           15:27:02
103 processes: 2 runnable, 93 sleeping, 8 on CPU
CPU0 states: 19.4% user,  0.0% nice, 77.6% system,  1.5% interrupt,  1.5% idle
CPU1 states: 22.9% user,  0.0% nice, 76.6% system,  0.0% interrupt,  0.5% idle
CPU2 states: 15.5% user,  0.0% nice, 78.5% system,  0.0% interrupt,  6.0% idle
CPU3 states: 14.9% user,  0.0% nice, 82.6% system,  0.0% interrupt,  2.5% idle
CPU4 states: 13.4% user,  0.0% nice, 84.1% system,  0.0% interrupt,  2.5% idle
CPU5 states: 14.9% user,  0.0% nice, 83.1% system,  0.0% interrupt,  2.0% idle
CPU6 states: 24.4% user,  0.0% nice, 75.1% system,  0.0% interrupt,  0.5% idle
CPU7 states: 55.2% user,  0.0% nice, 40.3% system,  0.0% interrupt,  4.5% idle
Memory: 3629M Act, 2046M Inact, 7572K Wired, 38M Exec, 5264M File, 518M Free
Swap: 9361M Total, 9361M Free
Seconds to delay: 
  PID USERNAME PRI NICE   SIZE   RES STATE      TIME   WCPU    CPU COMMAND
    0 root     124    0     0K  369M syncer/6  72:32  3.52%  3.52% [system]
19782 torrelay  26    0   310M  202M parked/5  19:39  2.78%  2.78% tor
25472 alnsn     77    0    11M 2336K select/6   0:00  3.88%  1.86% nbmake
 6642 alnsn     25    0    32M   16M CPU/1      0:00 33.00%  1.61% cc1
20956 alnsn     85    0  7508K 2296K select/4   0:51  0.63%  0.63% nbmake
10205 alnsn     77    0  5340K 1088K wait/5     0:00  1.26%  0.63% sh
 3554 alnsn     77    0  5700K 2464K select/7   0:00  1.38%  0.59% nbmake
 1191 alnsn     25    0    28M   11M CPU/7      0:00 11.00%  0.54% cc1
23337 alnsn     84    0    13M   11M kqueue/4  11:43  0.00%  0.00% tmux

-- 
Alex