[OpenAFS] Crash testing OpenAFS

ted creedon tcreedon@easystreet.com
Fri, 12 Aug 2005 12:53:55 -0700


Not good news:

cp -rvp /afs/.bigcell/foo/* /afs/.home.ted-doris.fam/bar #doesn't work

Where bigcell is 1.2.11 and home is 1.3.87

1. The transfer stops for no apparent reason after transferring for quite a
while. Looking at TOP - rxlistener, etc just disappear. The OS and afs
server/clients remained up.

2. Its doesn't appear to be the filenames. I built a replica of the
filesystem by making 1410 directories (with some weird names too) and
populating 24000 zero length files using touch.
This structure transferred in 38 sec.

3. It appears that the crashing is related to actual stress under load i.e.
when the 24000 files contain actual data.

4. cp -rvp /afs/.bigcell/foo/* /afs/.bigcell/bar #does seem to work.

5. I'll probably run tripwire on 4 this evening.

tedc

-----Original Message-----
From: openafs-info-admin@openafs.org [mailto:openafs-info-admin@openafs.org]
On Behalf Of ted creedon
Sent: Friday, August 12, 2005 8:46 AM
Cc: openafs-info@openafs.org
Subject: RE: [OpenAFS] Crash testing OpenAFS 

 

-----Original Message-----
From: openafs-info-admin@openafs.org [mailto:openafs-info-admin@openafs.org]
On Behalf Of chas williams - CONTRACTOR
Sent: Friday, August 12, 2005 3:23 AM
To: ted creedon
Cc: openafs-info@openafs.org
Subject: Re: [OpenAFS] Crash testing OpenAFS 

In message <20050812040746.EA4C9B0B8@smtpauth.easystreet.com>,"ted creedon"
writes:
>A simple cp -rpv /afs/.bigcell/foo /afs/.home-ted-doris.fam/bar hangs 
>the system so badly Linux won't even "halt".

that sounds simple, but what are foo and bar? 
>> foo= 4.8GB backup of a windows drive on a 1.2.11 AFS server bar=  
>> 8000000 Kblock mount into a volume on a 1.3.87 server

 i cant duplicate this test unless i have a little more information.

does afs break if you cp -rpv /afs/.bigcell/foo /afs/.bigcell/bar?
>>No. It runs fine if kept on the Linux 1.2.11 server/client

does it run for a little bit and stop? 
>>No. Takes a long while.
 does it hang immediately?
>>No
is there a particular file that stops it every time?
>> This is possibly the case. A month or 2 ago I dragged the same 
>> directory
from the 1.2.11 to a windows firewire drive using the Windows client and
observed duplicate filename messages from the windows boxes.

>>Jaltman mentioned that long filenames are not necessarily unique under
AFS, however they are unique in my 1.2.11 AFS filesystem, I don't know about
the 1.3.87 filesystem. I'll investigate.

_______________________________________________
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

_______________________________________________
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info