<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML dir=rtl><HEAD>
<META http-equiv=Content-Type content="text/html; charset=windows-1255">
<META content="MSHTML 6.00.6000.16609" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY dir=ltr bgColor=#ffffff>
<DIV><FONT face=Arial size=2><FONT size=3>No it just flashing the logs , the MDT
logs , im not a lustre expert but it fixed this issue when i run some testes
.</FONT></FONT></DIV>
<DIV><FONT face=Arial>Its also mentioned on some lustre wiki articles
.</FONT></DIV>
<DIV><FONT face=Arial></FONT> </DIV>
<DIV><FONT face=Arial size=2><FONT size=3></FONT></FONT> </DIV>
<DIV><FONT face=Arial size=2><FONT size=3></FONT></FONT> </DIV>
<DIV><FONT face=Arial size=2><FONT size=3></FONT></FONT> </DIV>
<DIV><FONT face=Arial size=2><FONT size=3>Are you saying to wipe the entire
cluster? That says it wipes the config logs for the fs, but I am not
entirely sure what that means...<BR><BR>--<BR>Andrew<BR><BR>> -----Original
Message-----<BR>> From: </FONT><A
href="mhtml:{00C4DF46-DB4E-424F-B56A-6BE88345A6E5}mid://00003763/!x-usc:mailto:lustre-discuss-bounces@lists.lustre.org"><FONT
size=3>lustre-discuss-bounces@lists.lustre.org</FONT></A><FONT size=3>
[mailto:lustre-discuss-<BR>> </FONT><A
href="mhtml:{00C4DF46-DB4E-424F-B56A-6BE88345A6E5}mid://00003763/!x-usc:mailto:bounces@lists.lustre.org"><FONT
size=3>bounces@lists.lustre.org</FONT></A><FONT size=3>] On Behalf Of Mailer
PH<BR>> Sent: Monday, March 17, 2008 12:07 PM<BR>> To: </FONT><A
href="mhtml:{00C4DF46-DB4E-424F-B56A-6BE88345A6E5}mid://00003763/!x-usc:mailto:Lustre-discuss@lists.lustre.org"><FONT
size=3>Lustre-discuss@lists.lustre.org</FONT></A><BR><FONT size=3>> Subject:
Re: [Lustre-discuss] how to replace a bad OST.<BR>><BR>> I run into
similar problem few weeks ago .<BR>><BR>> You need to run :<BR>>
tunefs.lustre --writeconf /dev/.............<BR>><BR>> On MDT/MGS after
unmounting it , maybbe there is another way to do that<BR>> without
unmounting MDT/MGS but im not sure .<BR>><BR>> Cheers
.<BR>><BR>><BR>><BR>> -----
Original Message -----<BR>> From:
Lundgren, Andrew<BR>> To: </FONT><A
href="mhtml:{00C4DF46-DB4E-424F-B56A-6BE88345A6E5}mid://00003763/!x-usc:mailto:'Lustre-discuss@clusterfs.com'"><FONT
size=3>'Lustre-discuss@clusterfs.com'</FONT></A><BR><FONT
size=3>> Sent: Monday, March 17, 2008
7:29 PM<BR>> Subject: [Lustre-discuss]
how to replace a bad
OST.<BR>><BR>><BR>> I am trying to
learn how to replace a defective OST with a new one.<BR>> Assuming the old
OST can not be
salvaged.<BR>><BR>><BR>><BR>> I
have a test cluster that I am working
on.<BR>><BR>><BR>><BR>> I
deactivated the volume on the MGS
using:<BR>><BR>><BR>><BR>> lctl
conf_param
content-OST0002-osc.osc.active=0<BR>><BR>><BR>><BR>>
I unlinked all of the bad files by finding the ones on the bad<BR>>
volume.<BR>><BR>><BR>><BR>> I
formatted a fresh OST using the index number of the bad
device:<BR>><BR>><BR>><BR>>
mkfs.lustre --reformat --fsname content --ost --<BR>> </FONT><A
href="mhtml:{00C4DF46-DB4E-424F-B56A-6BE88345A6E5}mid://00003763/!x-usc:mailto:mgsnode=4.248.52.81@tcp0"><FONT
size=3>mgsnode=4.248.52.81@tcp0</FONT></A><FONT size=3>
--param="failover.mode=failout" --index=02<BR>>
/dev/md6<BR>><BR>><BR>><BR>>
Then I tried to mount the freshly formatted OST into the
cluster.<BR>><BR>><BR>><BR>>
Unfortunately, I end up with an
error:<BR>><BR>><BR>><BR>>
mount.lustre: mount /dev/md6 at /lustre_raw_ost_one failed: Address<BR>>
already in use<BR>><BR>> The target
service's index is already in use.
(/dev/md6)<BR>><BR>><BR>><BR>>
How can I re-use the index number to prevent always having a "dead"<BR>>
point in my
cluster?<BR>><BR>><BR>><BR>>
Thanks!<BR>><BR>><BR>><BR>>
--<BR>><BR>>
Andrew<BR>><BR>><BR>><BR>>
----------------------------------------------------------<BR>> Outgoing
messages are virus free checked by NOD32 system</FONT><BR></FONT><FONT
face=Arial size=2><BR></DIV></FONT></BODY></HTML>