# ls -l /var/log/sysstat/ total 3296 -rw-r--r--. 1 root root 377076 Nov 4 00:00 sa03 -rw-r--r--. 1 root root 382268 Nov 5 00:00 sa04 -rw-r--r--. 1 root root 379672 Nov 6 00:00 sa05 -rw-r--r--. 1 root root 338184 Nov 7 00:00 sa06 -rw-r--r--. 1 root root 384 Nov 8 00:00 sa07 -rw-r--r--. 1 root root 384 Nov 9 00:00 sa08 -rw-r--r--. 1 root root 384 Nov 10 00:00 sa09 -rw-r--r--. 1 root root 384 Nov 11 00:00 sa10 -rw-r--r--. 1 root root 336 Nov 11 09:55 sa11 -rw-r--r--. 1 root root 381772 Nov 3 06:25 sar02 -rw-r--r--. 1 root root 376564 Nov 4 06:25 sar03 -rw-r--r--. 1 root root 381772 Nov 5 06:25 sar04 -rw-r--r--. 1 root root 379168 Nov 6 06:25 sar05 -rw-r--r--. 1 root root 338919 Nov 7 06:25 sar06 -rw-r--r--. 1 root root 0 Nov 8 06:25 sar07 -rw-r--r--. 1 root root 0 Nov 9 06:25 sar08 -rw-r--r--. 1 root root 0 Nov 11 06:25 sar10
The above is from one of my systems. It has had some reliability problems that caused it to crash but has now been up for 24 hours without problems, but I'm still getting the error messages. It seems that once I get this type of error it persists until I delete the sa files in question. I think that the system should be able to fix this. Obviously the program that is writing to the sa files is recognising that they are corrupt and skipping them. I think that it should be able to recognise that there is nothing useful and instead of just aborting it should truncate the file and write data as normal while logging an error message. -- My Main Blog http://etbe.coker.com.au/ My Documents Blog http://doc.coker.com.au/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
