daemon died for good after a loong query - help!
Posted in 1995
Topics: Storage & Space Management, Error Codes & Troubleshooting, Transactions, Locking & Isolation, Logging & Checkpoints
Hi,
It looks like our insert / select query crashed informix for good.
After we ran it, Informix crashed and we obtained a message like:
"Lock table overflow - userid107, ..
Program stopped at ...
SQL statement error number -271
Could not insert ...
SYSTEM error number -134
ISAM error - no more locks"
Nothing special, the _BIG_ problem is that since that time informix=20
wouldn't start.=20
When we reboot and try to start informix we get:
Tue Oct 3 08:16:20 1995=20
08:16:20 INFORMIX-OnLine Initialized -- Shared Memory Initialized=20
08:16:20 Physical Recovery Started
08:16:20 Checkpoint Completed
08:16:20 Physical Recovery Complete: 0 Pages Restored
08:16:20 ERROR: logundo(40) iserrno 126 us 0x101b2c pid 0=09tx 0x1028fc loguniq 1606 logpos 0x1d7260
08:16:20 INFORMIX-OnLine Must ABORT
=09Log Error 'rollback() - logundo() FAILED'=09us 0x101b2c pid 0 us_flags 0x
08:16:20 -- Online Aborting -- us=3D101b2c, pid=3D1366, uid=3D100
When we try tbinit -s it seems to run fine. But the daemon which tbinit=20
should start seems to die immediately - when we try to perform any=20
operation on the database (legal in quiescent mode) we get=20
=A8 daemon died " message.
Also, the user which ran the fatal query seems to be logged on forever.
Here's what we get when we run tbstat right after booting the machine and=
=20
running tbinit -s:
vic% tbstat -a
*** WARNING: INFORMIX-OnLine daemon is no longer running ***
RSAM Version 4.10.UC1 -- Abort -- Up 00:00:01 -- 1440 Kbytes
Message Log File: /usr/informix/turbo.log
tx 0x1028fc loguniq 1606 logpos 0x1d7260
13:47:23 INFORMIX-OnLine Must ABORT
Log Error 'rollback() - logundo() FAILED'
us 0x101b2c pid 0 us_flags 0x
13:47:23 -- Online Aborting -- us=3D101b2c, pid=3D1532, uid=3D100
Tue Oct 3 13:56:02 1995
13:56:02 INFORMIX-OnLine Initialized -- Shared Memory Initialized
13:56:02 Physical Recovery Started
13:56:02 Checkpoint Completed
13:56:02 Physical Recovery Complete: 0 Pages Restored
13:56:03 ERROR: logundo(40) iserrno 126 us 0x101b2c pid 0
tx 0x1028fc loguniq 1606 logpos 0x1d7260
13:56:03 INFORMIX-OnLine Must ABORT
Log Error 'rollback() - logundo() FAILED'
us 0x101b2c pid 0 us_flags 0x
13:56:03 -- Online Aborting -- us=3D101b2c, pid=3D1571, uid=3D100
Configuration File: /usr/informix/etc/tbconfig
=20
< I CUT OUT CONFIG FILE AS IT HAS ALWAYS WORKED FINE >
Users
address flags pid user tty wait tout locks nreads nwr=
ites
1019ac ------D 1571 informix ttyp2 0 0 0 6 2
101a0c ------D 0 informix ttyp2 0 0 0 0 0
101a6c ------F 1572 informix 0 0 0 0 0
101acc ------F 1573 informix 0 0 0 0 0
101b2c --R-X-- 0 rw 0 0 0 3 0
5 active, 32 total
Latches with lock or user set
name address lock wait user =20
Locks
address wtlist owner lklist type tblsnum rowid size =20
0 active, 1000 total, 64 hash buckets
Buffers
address user flgs pagenum memaddr nslots pgflgs xflgs owner waitl=
ist
0 modified, 500 total, 512 hash buckets, 2048 buffer size
Tblspaces
n address flgs ucnt tblnum physaddr npages nused npdata nrows nextns
0 10f16c 1 1 1000001 10000e 3995 633 0 0 1 =
=20
1 10f25c 1 1 1000260 10026d 992 988 1001 2958 9 =
=20
2 active, 320 total, 128 hash buckets
Dbspaces
address number flags fchunk nchunks flags owner name
1068ec 1 1 1 1 N informix rootdbs
1 active, 8 total
Chunks
address chk/dbs offset size free bpages flags pathname
105aac 1 1 0 199500 183116 PO- /dev/rsd2e
1 active, 24 total
Physical Logging
Buffer bufused bufsize numpages numwrits pages/io
P-2 0 16 0 0 0.00
phybegin physize phypos phyused %used =20
100fa8 500 285 0 0.00 =20
Logical Logging
Buffer bufused bufsize numrecs numpages numwrits recs/pages pages/io
L-1 0 16 0 0 0 0.0 0.0 =20
address number flags uniqid begin size used %used
121d6c 1 F------ 0 10119c 500 0 0.00
121d88 2 U---C-L 1606 101390 500 473 94.60
121da4 3 F------ 0 101584 500 0 0.00
121dc0 4 F------ 0 101778 500 0 0.00
Profile
dskreads pagreads bufreads %cached dskwrits pagwrits bufwrits %cached
11 29 16 31.25 4 5 2 0.00 =20
isamtot open start read write rewrite delete commit rol=
lbk
3 1 0 0 0 0 0 0 1
ovtbls ovlock ovuser ovbuff usercpu syscpu numckpts flushes=20
0 0 0 0 0.00 0.00 1 3 =20
bufwaits lokwaits lockreqs deadlks dltouts lchwaits ckpwaits compress
0 0 0 0 0 0 0 0 =20
*** WARNING: INFORMIX-OnLine daemon is no longer running ***
User rw seems to be logged on all the time.
We're running Informix OnLine 4.1 on SunOS 4.1.3
Help needed badly!
Thanks
Pawel Pietrusinski
///////////////////////////////////////////////////
///=09=09=09=09=09=09///
///Pawel Pietrusinski <pp@ciop.waw.pl>=09=09///
///=09=09=09=09=09=09///
///Central Institute for Labour Protection=09///
///Warsaw ul. Czerniakowska 16, =09=09///
/// Poland=09=09=09=09=09///
///////////////////////////////////////////////////=20
Related threads
- RE: transfer via comp.databases.informix
- dbimport hangs
- Assert Failed errno 271 ISAM ERR -12803
- Syslocks table