Hi Everyone,
I'm running II 9.0.4 (a64.lnx/105)NPTL +patch12125
I created a new database the other day, no problems with the create
but a subsequent checkpoint of the database (its first) gave the errors...
E_SC010D_DB_LOCATION Database Location Name: $default Physical
Specification: /dbdata1/II/ingres/data/default/ace_pack Flags: 00000003
E_DM0011_BAD_DB_NAME Illegal database name.
E_SC0121_DB_OPEN Error opening database. Name: ace_pack Owner:
ace Access Mode: 00000002 Flags 40000000
The checkpoint refused to die, I even tried removing the session from
the server but it refused to go.
In the end I had to crash the installation, but then the recovery server
refused to start, it gave a SIGSEGV in the DMF...
Segmentation Violation (SIGSEGV) @PC 00002b2dbeea74cd
RSP 00007fffffce7440 RBP 00007fffffce7620 RSI 00002b2dc6c6cd50
RDI 00002b2dc6c6cd50 RAX 0000000000000030 RBX 00002b2dc72d0138
RCX 00000000038c9c89 RDX 0000000000000018
Eventually I had to destroy and recreate my transaction logfiles to get
the damn thing to restart.
Clearly there is nothing wrong with the database name 'ace_pack'. Can
anyone shed some light on this?
Martin Bowes
--
Random Duckman Quote #117:
Duckman: The police ruled it was death by natural causes.
Cornfed: His body exploded into bits!
Duckman: Hey, the human body is an ever unfolding mystery.
martin.bowes@ctsu.ox.ac.uk - 26 Jul 2007 15:19 GMT
Hi Everyone,
Who wants to trash an installation? I've attached a test case for the
problem. You simply need a version of ingres which allows table
partitioning and a transaction log file of 1G or less...
Be warned that you will not clear the residual lock and problems
attaching to the database until the transaction log file is erased.
Failure to erase the log will result in a SIGSEGV from the the recovery
process when the installation restarts.
Execute test_case.sh <your database>
Marty
--
Random Farscape Quote #29:
Pilot - Luxan, you can't ambush me.
Ka'Dargo - I've got my gun pointed at your head! I think I'm doing pretty
well.