strange messages

Vernon Schryver vjs@calcite.rhyolite.com
Wed Aug 3 13:24:01 UTC 2005


> From: Pavel Urban 


> Aug  3 05:10:00 dcc1 dbclean[25406]: 1.3.3 cleaning dcc_db

there should have been preceding messages from dccd saying why it started
dbclean.  I suspect it was because of increase in the database size
due to the bug in 1.3.12.


> Aug  3 05:10:23 dcc1 dccd[900]: short packet of 1 bytes from 127.0.0.1,53986

do you have something that sends 1 byte packets to local sockets?


> Aug  3 05:36:04 dcc1 dccd[900]: sendto(OK, client 1 at 
> 81.164.109.16,40286: Resource temporarily unavailable
> Aug  3 05:39:39 dcc1 dccd[900]: sendto(OK, client 1 at 
> 24.16.115.230,37572: Resource temporarily unavailable

that is not good.
did the network system run out of mbufs or something?


> Aug  3 05:43:52 dcc1 dccd[900]: 1.3.3 database /var/dcc/dcc_db reopened 
> with 1018 MByte window

33 minutes is a long time for dbclean to run, but not unreasonable
for a public DCC server.

> Aug  3 05:43:52 dcc1 dccd[900]: "rejected ADMN FLOD CHECK request; 
> timestamp off by 33 seconds" sent to client 1030 at 127.0.0.1,53985
> Aug  3 05:43:52 dcc1 dbclean[25406]: ADMN FLOD CHECK: rejected ADMN FLOD 
> CHECK request; timestamp off by 33 seconds

are you using Linux?  That looks like a standard side effect of
how Linux and the last versions of BSDI's BSD/OS that I used
shuffle mmap() data between swap and the file system.


> I've never seen such logs before. Should I be worried?

I try to fix everything,
but those messages don't look fatal.


> I'm also seeing regular messages about excessive requests these days...

On a public DCC server, those can be stopped by " -R,,400" to DCCD_ARGS as in
    DCCD_ARGS="-u100,1000 -R,,400"
I do not know whether allowing 400 anonymous requests/second or up to
34 million/day is a good idea.  It is what I've been doing on
z.dcc-servers.net


Vernon Schryver    vjs@rhyolite.com



More information about the DCC mailing list

Contact vjs@rhyolite.com by mail or use the form.