[Pvfs2-developers] PVFS2 Error on startup

[David Metheny]

I ran the db_recovery tool, and by itself, that didn't fix enough to get the
server running.

 

I was going to make a backup copy of the data directory before trying other
things, but during the copy, got a bunch of Input/Output errors. That
triggered me to try an FSCK on the EXT3 file system I was mounting. After
that fsck (which found many problems), I tried starting the server again,
and it came back up. I in the process of seeing if all is there, but it's
running now, and files still appear to be in place. 

 

Thanks,

 

  _____  

From: Sam Lang [mailto:slang at mcs.anl.gov] 
Sent: Wednesday, May 27, 2009 4:39 PM
To: david.metheny at gmail.com
Cc: pvfs2-developers at beowulf-underground.org
Subject: Re: [Pvfs2-developers] PVFS2 Error on startup

 

 

On May 27, 2009, at 4:28 PM, David Metheny wrote:





We've got db4-4.2.52-7.3.el4 installed.

 

The value of ret after the line 650 call is '-30988'.

 

That's DB_PAGE_NOTFOUND.  You should try to recover the dspace.db file using
db_recover.

 

http://www.oracle.com/technology/documentation/berkeley-db/db/utility/db_rec
over.html

 

-sam





 

 

  _____  

From: Sam Lang [mailto:slang at mcs.anl.gov] 
Sent: Wednesday, May 27, 2009 11:27 AM
To: david.metheny at gmail.com
Cc: pvfs2-developers at beowulf-underground.org
Subject: Re: [Pvfs2-developers] PVFS2 Error on startup

 

 

That's not a very useful error, but it looks like there might be something
wrong with your dspace db.  Can you run it in gdb, set a breakpoint at
dbpf-dspace.c:650, step to the next line and print the result of "ret"?
Line 650 should have:

b dbpf-dspace.c:

 ret = dbc_p->c_get(dbc_p, &key, &data, DB_MULTIPLE_KEY|DB_NEXT);

 

The ret value is getting translated to a trove error code, except the BDB
error value isn't one that trove recognizes, so it just spits out an error
code that says it doesn't know anything about the error code.

 

What version of berkeley DB are you using?

 

-sam

 

On May 27, 2009, at 9:45 AM, David Metheny wrote:






I'm getting the following error starting up a PVFS2 server, running both IO
and metadata

 

PVFS2: [D] PVFS2 Server version 2.6.2pre1-2009-05-12-185735 starting.

PVFS2: [E] c_get failed on iteration 3059

PVFS2: [E] Warning: non PVFS2 error code (4243):

PVFS2: [E] dbpf_dspace_iterate_handles_op_svc: Unknown error 4243

PVFS2: [E] Error adding handle range
805306369-1073741823,2952790009-3221225463 to filesystem pvfs2-fs

PVFS2: [E] Error: Could not initialize server interfaces; aborting.

PVFS2: [E] Error: Could not initialize server; aborting.

 

The server is configured to sync both metadata and data. The server's
connected storage experienced a connection problem, causing a 'read-only
file system' error. After stopping and starting the server, it throws the
above errors each time.

 

I'm using the base 2.6 release, with many patches applied. Any ideas on how
to debug this, or recover from it?

 

Thanks,

 

 

_______________________________________________
Pvfs2-developers mailing list
Pvfs2-developers at beowulf-underground.org
http://www.beowulf-underground.org/mailman/listinfo/pvfs2-developers

 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.beowulf-underground.org/pipermail/pvfs2-developers/attachments/20090529/3652b6c3/attachment-0001.htm