[lustre-discuss] lnetctl net add does not alway work

[Thomas Roth]

Seems I found it:

First do "lctl network up", then do "lnetctl net add".

Of course the network is already up - how would my client use Lustre for 100s of days and 100s of TB without lnet? ;-)
But it is all started, say after a reboot, by mounting Lustre.  mount.lustre does load all modules and starts all processes, works fine for all but 
"lnetctl", which is capricious and wants the "network up" explicitly.


Cheers,
Thomas


On 23.06.2018 21:46, Thomas Roth wrote:
> Hi all,
> 
> 
> I have a test node and some active batch nodes, all running Lustre 2.10.3
> 
> They all have Lustre-1 mounted on the lnet "o2ib1" via "ib0".
> 
> 
> Now I try to add another net with
> 
> lnetctl net add --net o2ib5 --if ib0
> 
> This works on the test client but fails on the production nodes with
> 
> add:
>      - net:
>            errno: -22
>            descr: "cannot add network: Invalid argument"
> 
> 
> Other than the busy lnet on the production nodes I cannot see any difference. What did I miss?
> 
> 
> Regards,
> Thomas
> 
> 

-- 
--------------------------------------------------------------------
Thomas Roth
Department: HPC
Location: SB3 2.291
Phone: +49-6159-71 1453  Fax: +49-6159-71 2986

GSI Helmholtzzentrum für Schwerionenforschung GmbH
Planckstraße 1
64291 Darmstadt
www.gsi.de

Gesellschaft mit beschränkter Haftung
Sitz der Gesellschaft: Darmstadt
Handelsregister: Amtsgericht Darmstadt, HRB 1528

Geschäftsführung: Professor Dr. Paolo Giubellino
Ursula Weyrich
Jörg Blaurock

Vorsitzender des Aufsichtsrates: St Dr. Georg Schütte
Stellvertreter: Ministerialdirigent Dr. Rolf Bernhardt