<HTML>
<HEAD>
<TITLE>Re: [Lustre-discuss] Recovery fails if clients not connected</TITLE>
</HEAD>
<BODY>
<FONT FACE="Verdana, Helvetica, Arial"><SPAN STYLE='font-size:12.0px'><BR>
Hi Roger,<BR>
<BR>
I believe you can connect the OSSs once the MDS has booted, and in fact, I’m pretty sure that the five in the ‘connected_clients: 0/5’ are in fact your OSS nodes. Each OST maintains a connection to the MDS while the file system is mounted, so they will be included in the connection count on the MDS.<BR>
<BR>
However, regardless of the state — if your MDS is online and the MDT is mounted, you can start up the OSS nodes and corresponding OSTs at any time; clients attempting to make transactions will have their I/O operations block (or fail, depending on the MDS config) until the missing nodes come back online.<BR>
<BR>
hth,<BR>
Klaus<BR>
<BR>
<BR>
On 1/20/09 3:05 PM, "Roger Spellman" <roger@terascala.com> etched on stone tablets:<BR>
<BR>
</SPAN></FONT><BLOCKQUOTE><FONT SIZE="4"><FONT FACE="Courier New"><SPAN STYLE='font-size:13.0px'>I have 2 MDS, configured as an active/standby pair. I have 5 OSTs that are NOT active/standby. I<BR>
have 5 clients.<BR>
<BR>
I am using Lustre 1.6.5, due to bug 18232 <https://bugzilla.lustre.org/show_bug.cgi?id=18232> which only affects 1.6.6. Using Lustre 1.6.5, when I<BR>
reset my active node, the standby takes over. This is quite reliable.<BR>
<BR>
Today, I did the following in this order:<BR>
Unmounted all the clients<BR>
Rebooted all the clients<BR>
Stopped Linux HA from running<BR>
Unmounted the OSTs<BR>
Unmounted the MDS<BR>
Rebooted the OSTs<BR>
Rebooted both MDSes<BR>
<BR>
When the MDSes started up, Linux HA chose one to be active. That system mounted the MDT.<BR>
<BR>
I looked at the file /proc/fs/lustre/mds/tacc-MDT0000/recovery_status, and it showed:<BR>
<BR>
[root@ts-tacc-01 ~]# cat /proc/fs/lustre/mds/tacc-MDT0000/recovery_status <BR>
status: RECOVERING<BR>
recovery_start: 0<BR>
time_remaining: 0<BR>
connected_clients: 0/5<BR>
completed_clients: 0/5<BR>
replayed_requests: 0/??<BR>
queued_requests: 0<BR>
next_transno: 17768<BR>
<BR>
<BR>
***** Note that recovery_start and time_remaining are both zero. *****<BR>
<BR>
I waited a several minutes, and this file was the same.<BR>
<BR>
I was waiting for recovery to complete before trying to mount the OSTs. However, it appears that<BR>
this would never occur!<BR>
<BR>
Does this look like a bug? <BR>
<BR>
---------------------------<BR>
<BR>
I format my MDT using the following command. The command is run from 10.2.43.1, and the failnode<BR>
is 10.2.43.2:<BR>
<BR>
mkfs.lustre --reformat --fsname tacc --mdt --mgs --device-size=10000000 --mkfsoptions=' -m 0 -O<BR>
mmp' --failnode=10.2.43.2@o2ib0 /dev/sdb<BR>
<BR>
I format the OSTs using the following command:<BR>
<BR>
/usr/bin/time -p mkfs.lustre --reformat --ost --mkfsoptions='-J device=/dev/sdc1 -m 0' --fsname<BR>
tacc --device-size=400000000 --mgsnode=10.2.43.1@o2ib0 --mgsnode=10.2.43.2@o2ib0 /dev/sdb<BR>
<BR>
I mount the clients using:<BR>
<BR>
mount -t lustre 10.2.43.1@o2ib:10.2.43.2@o2ib:/tacc /mnt/lustre<BR>
</SPAN></FONT><SPAN STYLE='font-size:13.0px'><FONT FACE="Arial"> <BR>
</FONT></SPAN></FONT><FONT FACE="Verdana, Helvetica, Arial"><SPAN STYLE='font-size:12.0px'><BR>
<HR ALIGN=CENTER SIZE="3" WIDTH="95%"></SPAN></FONT><FONT SIZE="2"><FONT FACE="Monaco, Courier New"><SPAN STYLE='font-size:10.0px'>_______________________________________________<BR>
Lustre-discuss mailing list<BR>
Lustre-discuss@lists.lustre.org<BR>
<a href="http://lists.lustre.org/mailman/listinfo/lustre-discuss">http://lists.lustre.org/mailman/listinfo/lustre-discuss</a><BR>
</SPAN></FONT></FONT></BLOCKQUOTE><FONT SIZE="2"><FONT FACE="Monaco, Courier New"><SPAN STYLE='font-size:10.0px'><BR>
</SPAN></FONT></FONT>
</BODY>
</HTML>