Re: more bgpd weirdness

[Arvid Grøtting <arvidg_(_at_)_netfonds_(_dot_)_no>]

Henning Brauer <lists-openbsd_(_at_)_bsws_(_dot_)_de> writes:

> a high load average is nothing to worry about, however it seems you got 
> very high CPU use too, which does not match at all with our tests and 
> our priduction uses. so something is strange with your box/setup/... 
> and we have to find out what - and as we have seen quite some bad 
> builds....

By bad builds, you mean what, exactly?

Anyway, scp-ing over src.tar.gz and such, I noticed that things didn't
go very smoothly there, either.  Here's top, on the bad box (dmesg
sent yesterday), while scp-ing src.tar.gz and sys.tar.gz at
pain-stakingly slow (for 100Mb/s media) 229.3KB/s:

load averages:  1.17,  1.11,  0.79                                     15:30:03
28 processes:  1 running, 25 idle, 1 stopped
CPU states:  3.7% user,  0.0% nice,  2.2% system, 94.1% interrupt,  0.0% idle
Memory: 40M/105M act/tot  Free: 386M  Swap: 0K/1024M used/tot

  PID USERNAME PRI NICE  SIZE   RES STATE    WAIT     TIME    CPU COMMAND
 1624 arvidg    56    0  528K 1316K run      -        0:42 87.26% sshd
27800 arvidg    -6    0  124K  476K sleep    piperd   0:00  0.05% scp
[...]

And a bit later, same transfer:

load averages:  2.09,  1.83,  1.24                                     15:35:41
28 processes:  1 running, 25 idle, 1 stopped
CPU states:  2.5% user,  0.0% nice,  0.8% system, 96.8% interrupt,  0.0% idle
Memory: 40M/105M act/tot  Free: 386M  Swap: 0K/1024M used/tot

  PID USERNAME PRI NICE  SIZE   RES STATE    WAIT     TIME    CPU COMMAND
 1624 arvidg    64    0  628K 1428K run      -        6:16 100.54% sshd
 2188 root       2    0  968K   32M sleep    select  10:55  0.00% gated
[...]

This does not look normal.  Not at all.

After the transfer's over, everything's fine again:

load averages:  0.86,  1.60,  1.26                                     15:38:26
24 processes:  22 idle, 1 stopped
CPU states:  0.0% user,  0.0% nice,  0.0% system,  2.8% interrupt, 97.2% idle
Memory: 39M/104M act/tot  Free: 387M  Swap: 0K/1024M used/tot

  PID USERNAME PRI NICE  SIZE   RES STATE    WAIT     TIME    CPU COMMAND
 2188 root       2    0  968K   32M sleep    select  10:55  0.00% gated
28078 _pflogd    4    0  444K  200K sleep    bpf      6:20  0.00% pflogd

I've tested scp-ing it to /dev/null, and that didn't work any faster,
so I don't think it's the disk slowing me down.

I'm suspecting ste(4) and the D-Link 580 card here, and I'll try
swapping it for a spare 570 shortly.  Stay tuned.


-- 

Arvid