System crashes every 3 weeks.

[John Travell]

Doug,

If you can FTP copy a valid crash dump to my server both Volker and I will be able to look through it and see if anything can be found.
Look up my profile and send me an email for the ftp login details.

John Travell.

If we get some results we can discuss whether the answers are worth anything. Possibly a suggestion to your TPM that there are independent backup resources available.. :-)

[Volker Halle]

Doug,

if we can confirm, that no mchk logout frame has been logged by the hardware, then the next step would be to try to figure out (from the instruction stream data calling EXE$GEN_BUGCHK) which code has actually called the MACHINECHK bugcheck. This may be quite hard, since this is an old (unsupported) version of OpenVMS.

The other thing would be to try and figure out, what kind of IO-operation was going on at the time of the crash. You can look at busy devices:

SDA> SET OUT x.x
SDA> SHOW DEV
SDA> EXIT
$ SEA x.x bsy

Volker.

[Volker Halle]

Doug,

there is still some time left until the next crash...

Could you please issue the following commands against the dump:

$ ANAL/CRAS sysdump.dmp
SDA> READ/EXEC
SDA> EXA/INS 80011C80:80011E04+10

This will show the V7.1-1H2 instruction stream from the ERROR_ROUTINES_1102 and I may be able to find out, exactly which MACHINECHK in this module has been taken. This is to better understand the crash context.

Volker.