How can I check when my system last crashed?
Or how do I know that the dump isn’t “stale”?
You can check the date/timestamp of the most recent crash dump contained in the system’s dump file. This can be done both with the “active” dump file and with any copy of it.
Use ANALYZE /CRASH_DUMP
to check the “Dump taken on
” and the “Time of system crash:
” date/timestamps.
Note that, other than read-access to the Dump File itself, there are no special privileges needed to perform this particular operation.
%PRIVILEGE-I-ENABLED: TMPMBX,OPER,NETMBX
$ ANALYZE /CRASH sys$system:sysdump.dmp
OpenVMS system dump analyzer ...analyzing an I64 compressed selective memory dump... %SDA-W-LMBMISSING, "Key Global Pages" LMB missing from dump Dump taken on 16-NOV-2016 01:49:16.43 using version V8.4 OPERCRASH, Operator forced system crash
SDA> SHOW CRASH
System crash information ------------------------ Time of system crash: 16-NOV-2016 01:49:16.43 Version of system: OpenVMS I64 Operating System, Version V8.4 System Version Major ID/Minor ID: 3/0 VMScluster node: CLASS8 System type: HP rx2600 (1.40GHz/1.5MB) CPUs are not thread-capable Primary CPU ID: 000 (0.) Crash CPU ID: 000 (0.) Bitmask of active CPUs: 00000000.00000003 Bitmask of available CPUs: 00000000.00000003
Other information in this check includes the version of VMS, the system’s nodename, the system type (model), and the number of CPUs in the system.
If the date/timestamp that you see herein does not correspond to your own (or others’) knowledge of when the system crashed, it’s likely that you are looking at a “stale” (obsolete) crash dump, and that the most recent crash actually did not dump memory to the dump file.
If this is the case, you'll need to look at and likely correct certain VMS crash dump parameters and configuration to capture a good, complete crash dump for analysis. Work with your PARSEC Tech Support specialist to determine how to fix this configuration for effective and complete crash dumps.