000019694 - ACE/Server does not start after a hardware failure

Document created by RSA Customer Support Employee on Jun 14, 2016Last modified by RSA Customer Support Employee on Apr 21, 2017
Version 2Show Document
  • View in full screen mode

Article Content

Article Number000019694
Applies ToRSA ACE/Server 5.0.1 (no longer supported as of 8-15-2004)
UNIX (AIX, HP-UX, Solaris)
IssueACE/Server does not start after a hardware failure
Replaced faulty hardware but unable to start ACE/Server database broker
Error: "rlrdnxt: note prefix and suffix lengths don't match 552 9"
The following messages appear when ACE/Server is started:
Message: Starting server database broker and page writers.
12:08:32 BROKER  0: Multi-user session begin. (333)
12:08:32 BROKER  0: ** The last session was abnormally terminated. (852)
12:08:32 BROKER  0: ** Any incomplete transactions are being backed out. (459)
12:08:32 BROKER  0:  rlrdnxt: note prefix and suffix lengths don't match 552 9
12:08:32 BROKER  0: SYSTEM ERROR: The broker is exiting unexpectedly, beginning
Abnormal Shutdown. (5292)
12:08:32 BROKER  0: drexit: Initiating Abnormal Shutdown
12:08:32 BROKER  0: ** Save file named core for analysis by Progress Software
Corporation. (439)
12:08:32 BROKER  0: SYSTEM ERROR: Releasing regular latch. latchId: 4 (5028)
12:08:32 BROKER  0: User 0 died holding 1 shared memory locks. (2522)
12:08:32 SERVER   : ** The server terminated due to signal 3. (801)
ResolutionTo correct this issue, the database must be restored from a backup tape.
The following information is based on the Progress Software Technical Support Note #17247. This is a corruption error of the notes in the before image file. These notes, kept in order with transaction numbers, track all changes to your database. The before image file is used at database startup and during crash recovery to reconcile any changes to the database. This file is crucial to ensuring the integrity of your data.

Corruption to the before image file is usually caused by hardware errors or errors at the operating system level. Unless you correct this issue, these problems are likely to reoccur.

If you get this error, restore your backup and roll forward your after image files (if you are using after imaging).

If for some reason this is not a feasible resolution (i.e. your backup quality is suspect), you can force truncate the before image file. This action may cause logical corruption in your database, and your database must be repaired (dump and load of database is the recommended repair procedure), but it will allow access to the database after the truncate.

The syntax to force truncating your before image file is as follows:

        proutil dbname -C truncate -F    (-F is the force flag)

You will then see the following messages:

** The FORCE option was given, database recovery will be skipped. (33)
After-imaging disabled by the FORCE parameter. (4790)
** Your database was damaged. Dump its data and reload it. (37)

You will see the (37) error every time you start this database until the repair on the database is completed. It is recommended you repair your database as soon as possible after forcing in.

Please see the following notes from Progress Software's Support Knowledgebase for additional information:

Technical Support Note #14115 The Perils of using the Force ( -F ) option to start the db
Technical Support Note #16774 Force Access -F Parameter in Version 8.2 and later
Legacy Article IDa11466