2903224 - Error 632 may be raised during shrinking a database with many device fragments - SAP ASE

• A 632 error may be raised when trying to shrink a database by ALTER DATABASE OFF/LOG OFF command with a device name and the database has multiple device fragments of the specified device name.
• For example :

1> alter database <database name> log off <device name>
2> go
Msg 632, Level 20, State 2:
Server 'ASE_name', Line 1:
Attempt to move memory with an incorrect length of 18432. Maximum allowed length is 16384.

• With or without a 632 error, a "Current process infected with 11 (SIGSEGV)" error or system exception (0xc0000005) may follow during cleaning up.
  The stack trace in ASE errorlog looks like below (some functions might be missing) :

altdb__free_device_fragments
altdb__cleanup_auhdr
altdb__shrinklog_cleanup
altdb__shrinklog_backout
exc__raise
exc_raise
memmove_error
altdb__realloc_auhdr
altdb__track_mod_req
altdb_shrinklog
alterdb

• Adaptive Server may shut down with the message "FATAL UNHANDLED EXCEPTION: signal 0 hit while handling a previously hit signal. The server cannot continue and will shut down.".

• SAP Adaptive Server Enterprise 16.0
• SAP Adaptive Server Enterprise 16.0 for Business Suite

CR#821254, CR821254, #821254, altdb__unlocklog, xls_unlocklog, sysusages , KBA , BC-SYB-ASE , Sybase ASE Database Platform (non Business Suite) , BC-DB-SYB , Business Suite on Adaptive Server Enterprise , Bug Filed