Helix Debian Packages and a weird error
OK. Weird error here, and I'm at a loss on how to track it down. I'm
sending you this email because you're one of the Debian package
maintainers named "Peter". I realize how thin this lead is, however.
Anyway, on to the error...
My home directories are automatically mounted via NIS and NFS with the
autofs package (automount utility). Periodically, this error pops up
in my syslog:
-------[ SYSLOG SNIP ]------------------------------------------------
Sep 29 15:26:20 ws96-8 automount[251]: attempting to mount entry /home/peter
Sep 29 15:26:20 ws96-8 automount[14346]: lookup(yp): looking up peter
Sep 29 15:26:20 ws96-8 automount[14346]: lookup(yp): peter -> -rw,soft^Ishepard:/export/home/&
Sep 29 15:26:20 ws96-8 automount[14346]: expanded entry: -rw,soft^Ishepard:/export/home/peter
Sep 29 15:26:20 ws96-8 automount[14346]: parse(sun): gathered options: rw,soft
Sep 29 15:26:20 ws96-8 automount[14346]: parse(sun): core of entry: shepard:/export/home/peter
Sep 29 15:26:20 ws96-8 automount[14346]: mount(nfs): calling mkdir /home/peter
Sep 29 15:26:20 ws96-8 automount[14346]: mount(nfs): calling mount -t nfs -s -o rw,soft shepard:/export/home/peter /home/peter
Sep 29 15:26:20 ws96-8 automount[14346]: >> mount: shepard:/export/home/peter failed, reason given by server: No such file or directory
Sep 29 15:26:20 ws96-8 automount[14346]: mount(nfs): nfs: mount failure shepard:/export/home/peter on /home/peter
----------------------------------------------------------------------
This is only one block of what amounts to spam in my syslog as some
application tries to browse the home directory of "peter". Since it
does not find peter as a local user (my fix, by the way), it goes to
the NIS server. There IS an entry for peter in the auto.passwd file,
but there is no export home directory on the NFS server for said
peter.
I haven't checked the frequency of this occurrance, my next step in
debugging this, but this is a very consistent and very rapid activity.
Much too rapid for any human trying to log in, lasting for up to two
minutes. (Throwing hundreds of records into syslog of the manner
above.)
OH! Hey! I found another attempt right after gconfd fired up while
running nautalis. Then I found an attempt for this one...
----------------------------------------------------------------------
Sep 29 16:47:48 ws96-8 automount[15108]: mount(nfs): calling mount -t nfs -s -o rw,soft shepard:/export/home/.nautilus-metafile.xml /home/.nautilus-metafile.xml
Sep 29 16:47:48 ws96-8 automount[15108]: >> mount: shepard:/export/home/.nautilus-metafile.xml failed, reason given by server: No such file or directory
----------------------------------------------------------------------
Are we getting somewhere here?! Woohoo! So, it looks like the
problem lies with the Helix applications that are associated with
gconfd. Have you programmed your name in as a default directory for
gconfd to look into? (Assuming you're a contributor to the helix
project).
Doh. Well, automount *is* still trying to mount the nautilus
metafile, but I can't get it to try to mount "peter" any more (well, I
did add a local user named "peter" in home.local). Well, let me know
if you come up with anything. I'm going to forward this to the Debian
User list as well to see if anyone else is having this problem.
I'll run some more tests w/o the local "peter" account and see if I
can associate with gconfd.
--
Chad "^chewie" Walstrom, DI
mailto:chad.walstrom@iris.ebenx.com
Reply to: