[lustre-discuss] mounting OSTs of failed node on other nodes

E.S. Rosenberg esr+lustre at mail.hebrew.edu
Tue Nov 28 16:51:48 PST 2017

So it turns out that when I setup this lustre filesystem I defined it for
failover even though the vendor back then (2.4-2.5 days) was very negative
about it.
We never had an opportunity to actually test failover properly but it
worked nicely now, thanks to all the great people who work on this project!

I am still curious though, assuming I had not done the original setup for
failover what steps would I now take to make these OSTs available?
I assume it's something like:
- mount OSTs
- tune2fs --writeconf [what?]


Thanks again!

On Wed, Nov 29, 2017 at 2:40 AM, Brian Andrus <toomuchit at gmail.com> wrote:

> It should be possible.
> You will need to update the config (writeconf) to reflect the IP of the
> host (failnode or servicenode).
> This way, when it reports to the MGS, it is known and will be accepted
> into the mix :)
> Brian Andrus
> On 11/28/2017 4:35 PM, E.S. Rosenberg wrote:
> Hi everyone,
> One of our OSS nodes failed to come back up after a firmware upgrade.
> I'd like to mount its' OSTs on one or more of the other OSSs to allow the
> filesystem to remain available.
> Is this possible?
> Thanks,
> Eli
> _______________________________________________
> lustre-discuss mailing listlustre-discuss at lists.lustre.orghttp://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
> _______________________________________________
> lustre-discuss mailing list
> lustre-discuss at lists.lustre.org
> http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lustre.org/pipermail/lustre-discuss-lustre.org/attachments/20171129/eb1c5ee0/attachment-0001.html>

More information about the lustre-discuss mailing list