[454] in Moira
Re: Unknown code sms 255
daemon@ATHENA.MIT.EDU (Jim Haynes)
Mon Sep 14 20:42:14 1992
From: haynes@cats.UCSC.EDU (Jim Haynes)
Date: Mon, 14 Sep 92 11:30:01 -0700
To: bug-moira@Athena.MIT.EDU
Interestingly enough, we had the same error on the same day
Sep 11 14:12:17 <1275> dcm: checking prlogger...
Sep 11 14:12:19 <1275> dcm: running /u1/sms/bin/prlogger.gen
Sep 11 14:12:22 <1275>
Sep 11 14:12:22 <1275> SYSERR: SET_LOCK: err:cbe9c7
Sep 11 14:12:22 <1275>
Sep 11 14:12:22 <1275>
Sep 11 14:12:22 <1275>
Sep 11 14:12:22 <1275> PROGRAM ERROR: 1008 Illegal context of "retrieve(flag1=int4(interval(""
Sep 11 14:12:22 <1275> without preceding ## INGRES statement. Exiting...
Sep 11 14:12:22 <1275> dcm: Unknown code sms 255 /u1/sms/bin/prlogger.gen exitedSep 11 14:12:23 <1275> dcm: DCM building config files for prlogger: Unknown code sms 255
-----
Later (swirl of electronic organ music)...
Sep 12 03:42:05 <4284> dcm: checking hesiod...
Sep 12 03:42:05 <4284> DIaddpage: file = 18 io=cef2c pagenum=0 written=-1
Sep 12 03:42:05 <4284>
Sep 12 03:42:05 <4284>
Sep 12 03:42:05 <4284> Run restoredb after correcting the problem
Sep 12 03:42:05 <4284> Unable to read from program: Interrupted system call
Sep 12 03:42:05 <4284> exited with code 255
Sep 12 03:57:01 <4296> dcm: Ingres system error occurred code 19
Sep 12 03:57:01 <4296>
Sep 12 03:57:01 <4296> dcm: Moira server encountered INGRES ERROR 19
Sep 12 03:57:01 <4296> INGRES ERROR: 19 Request for a lock failed.
Sep 12 03:57:01 <4296> Unable to read from program: Interrupted system call
-----
I wasn't here the 11th, and the guy who was is now away on vacation, but the
moira.log shows the moirad might have been down on Sep 11 at 14:12, as it
was restarted later. Here's moira.log excerpt.
Sep 11 13:17:25 <165> moirad[#54]: Query complete.
Sep 11 13:17:32 <165> moirad[#54]: Closed connection (now 0 clients, 200 new queries, 0 old)
Sep 11 15:20:09 <165> moirad: started (pid 165)
Sep 11 15:20:09 <165> moirad: $Header: /afs/athena.mit.edu/astaff/project/moiradev/src/server/RCS/mr_main.c,v 1.29 90/06/13 12:44:14 mar Exp $
----
Now early the next morning...
Sep 12 05:40:24 <165> moirad[#3]: New connection from 128.114.129.26 port 3458 (now 1 client)
Sep 12 05:40:24 <165> moirad[#3]: motd[2]:
Sep 12 05:40:24 <165> moirad[#3]: auth[2]: "\004\006\002CATS.UCSC.EDU", "chfn"
Sep 12 05:40:26 <165>
Sep 12 05:40:26 <165> SYSERR: The database sms may be inconsistent. Please run RESTOREDB.
Sep 12 05:40:26 <165>
Sep 12 05:40:26 <165>
Sep 12 05:40:26 <165> moirad[#3]: 174: child exits with signal 0 status 255
Sep 12 05:40:26 <165> Unable to read from program: Interrupted system call
Sep 12 05:40:26 <165> exited with code 255
----
I noticed all this that night, and the moira daemon was not running, so I
ran restoredb and then dbck and restarted the daemon and all has been well
since then. Sure is curious that we both had the same problem on opposite
coasts on the same day - makes you wonder if there is some calendar-related
bug deep within Ingres. The Ingres we are running is
INGRES SunOS Release 5.0/04a (sun.u42/07)
Our prlogger is not quite the same as yours - because of differences in the
data base we can't use your SQL-based ModDiff() function, so I hacked up
one of the other QUEL-based files to do approximately what that function
does and used that as a substitute ModDiff.
