Memory fault(coredump) error when using 'last' command

[Simon R Wootton]

If I issue a 'last' command

eg. last root

I get a 'Memory fault(coredump)' message. Has anyone seen this before ? Any idea's on cause/rectification ? Could it be to do with the wtmp file ?

ALl help rewarded.
Thanks
Simon

[Pete Randall · Accepted solution]

Simon,

The usual culprit is a corrupt wtmp file. Try moving it out of the way and re-creating it with "cat /dev/null > /var/adm/wtmp" to see if that cures the problem.


Pete

[James R. Ferguson]

Hi Simon:

Yes, it is highly likely that 'var/adm/wtmp' (or perhaps 'var/adm/btmp') is corrupt. You can preserve (copy) both and recreate each to see if the problem disappears. If you find 'wtmp' causes the coredump but not 'btmp' you can restore 'btmp' and null 'wtmp'. To recreate, do:

# cat /dev/null > /var/adm/wtmp
# cat /dev/null > /var/adm/btmp

Regards!

...JRF...

[Simon R Wootton]

Problem sorted - thanks guys !

[Tim Adamson_1]

Hi,

It sounds like you may have a corrupt wtmp file or btmp.

Try copying them aside for backup purposes, then zero one out (eg, > wtmp) and try the last command again. If it works then you have found the problem. If not, then zero the other one out and cp the first one back and try again.

Cheers!

[Steve Steel]

Hi

99 percent chance of wtmp problems

look at wtmpfix

Steve Steel

[Jim Mills_3]

I had the same problem and just nulled the wtmp file and it fixed my problem. Thanks.