[Sisuite-users] mke2fs of non-mounted filesystems

2004-07-20 Thread Dan
Am I correct in thinking that when getimage takes an image of the golden
client, it will generate parted commands for non-mounted filesystems, but not
generate an mke2fs command for that filesystem?

I just ask because our golden client has /dev/hdb1 which is unmounted most of
the time, and remounted at midnight to accept dumps of /dev/hda* partitions.  

We've started to experience installations which fail on reboot post
installation, complaining about "bad magic number block on /dev/hdb1".

We have been fixing this by having users enter the root password and
fdisk/mke2fs /dev/hdb1 (over the phone, with non-Linux savvy users this has
been fun) whilst we've tried to figure out what was wrong.  

Today we were poring over things trying to figure it out once and for all and
closer examination shows us that the .master file has parted entries for
/dev/hda* and /dev/hdb1, but only has mke2fs/mount commands for /dev/hda*.

Is this a feature?  

cheers!

Dan





___ALL-NEW Yahoo! Messenger - 
so many all-new ways to express yourself http://uk.messenger.yahoo.com


---
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=4721&alloc_id=10040&op=click
___
Sisuite-users mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/sisuite-users


Re: [Systemimager] Re: [Sisuite-users] Flamethrower connectivity test problem

2004-07-20 Thread Brian Elliott Finley
Ah!
It just hit me.  The reason we don't ping the IMAGESERVER if
FLAMETHROWER_DIRECTORY_PORTBASE is set, is that the client may never be
given, know, or need to know, the ip address of the imageserver because
the client is receiving _all_ of it's data via multicast, which is more 
like listening to a channel, as compared with connecting directly to 
a server. 

I've put this as a comment in there for a future release.
Cheers, -Brian
Thus spake Bas van der Vlies ([EMAIL PROTECTED]):
Brian Elliott Finley wrote:
Thus spake Bas van der Vlies ([EMAIL PROTECTED]):
We at SARA are currently switching to flamethrower setup. If  we 
enable flamethrower the conenctivity test is different then
with rsync, from /etc/init.d/rcS

#   Ping test
ping_test() {
echo
echo ping_test
if [ ! -z "$FLAMETHROWER_DIRECTORY_PORTBASE" ]; then
PING_DESTINATION=$GATEWAY
HOST_TYPE="default gateway"
else
PING_DESTINATION=$IMAGESERVER
fi
echo
echo "Pinging your $HOST_TYPE to ensure we have network 
connectivity."
echo

My question is why?

I honestly don't remember why, but it definitely looks like it was on
purpose.

It is fine for me ;-). Still strange that the check is different. Maybe 
we have to do 2 ping tests if flamethrower is enabled. One to the 
gateway and if this fails to the imageserver. Just a thought

Regards
--
--

*  *
*  Bas van der Vlies e-mail: [EMAIL PROTECTED]  *
*  SARA - Academic Computing Servicesphone:  +31 20 592 8012   *
*  Kruislaan 415 fax:+31 20 6683167*
*  1098 SJ Amsterdam   *
*  *

--
--
Brian Elliott Finley   Argonne, MCS Division 
Mobile: 630.631.6621   Office:  630.252.4742
gpg --keyserver wwwkeys.pgp.net --recv-keys 10F8EE52
--

---
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=4721&alloc_id=10040&op=click
___
Sisuite-users mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/sisuite-users