Hi all;
I've got a weird one going on here. I ran a 3 days without tape
test to check amflush after Jon had a problem, and it worked just
fine, flushing to a total of 4 tapes to do it all.
Now I want to repeat that test, but this time I was going to let the
autoflush option handle it. Buit I now have a connection problem!
So it didn't do any backups this morning.
I'd been adding to the hosts file, two more lines plumb out of the
localdomain, but I don't think those addresses should have had
anything to do with amanda localhost connection done through a FQDN
in the hosts file.
I'd heard that the hosts file should be totally bereft of any
terminating linefeeds, so I just took 2 of them out, amcheck still
reports;
WARNING: coyote: selfcheck request timed out. Host down?
Also, and I find this a bit odd, xinetd keeps running amandad,
restarting it if I kill it in about 5 seconds. AFAICS there is no
caller that should make amandad run, and kpm shows the parent
process is xinetd.
And I just found a mutt zombie with kpm and killed it as it was
eating 80% of the cpu, no idea how long its been hanging around,
but that didn't fix it, not that I expected it to. One more
upchuck out of mutt and its gonna be rpm -e.
I just made sure the .amandahosts file was valid, and had 0600 for
perms.
Anybody have any clues? Humm, before mailing this, I took those two
extra names out of the hosts file, and now amcheck works. And I
assume if I fired off an amdump session, it would too.
Can anyone explain that? Why should adding the adresses of the mail
servers at my SP and my workplace to the hosts file cause amdump
and amcheck to fail in the above manner?
Weird, and no connection to Weird Al Yankovic as I can't hear any
music at all right now. Snilmerg maybe?
--
Cheers, Gene
AMD K6-III@500mhz 320M
Athlon1600XP@1400mhz 512M
99.22% setiathome rank, not too shabby for a WV hillbilly
|