[lustre-discuss] common error during periods of slow client iO

[Dilger, Andreas]

On Oct 8, 2017, at 18:51, John White <jwhite at lbl.gov> wrote:
> 
> We get this error quite often when OSSs are providing very slow client IO and I'd love to know what it means.  I'd imagine it's one of those situation where Andreas says, "this is a symptom, not a cause"

Sorry, I'm not sure what this is.  At first glance I'd say some kind of network
error is causing problems during data transfer.  The "-125" error (-ECANCELED)
is returned in a few different parts in the code, but is most likely from
the LNet layer. 

> LustreError: 4043:0:(events.c:452:server_bulk_callback()) event type 5, status -125, desc ffff8803419d5000
> LustreError: 4041:0:(events.c:452:server_bulk_callback()) event type 5, status -125, desc ffff880174d7e400
> LustreError: 4041:0:(events.c:452:server_bulk_callback()) event type 5, status -125, desc ffff8806c5f32600
> LustreError: 4041:0:(events.c:452:server_bulk_callback()) event type 5, status -125, desc ffff8806c6717a00
> LustreError: 4041:0:(events.c:452:server_bulk_callback()) event type 5, status -125, desc ffff880297d33800
> LustreError: 4041:0:(events.c:452:server_bulk_callback()) event type 5, status -125, desc ffff880155e58000
> LustreError: 4044:0:(events.c:452:server_bulk_callback()) event type 5, status -5, desc ffff8806a8d33200
> LustreError: 4044:0:(events.c:452:server_bulk_callback()) event type 5, status -125, desc ffff8806c6715a00
> LustreError: 4044:0:(events.c:452:server_bulk_callback()) event type 5, status -5, desc ffff8801fd15cc00
> LustreError: 4044:0:(events.c:452:server_bulk_callback()) event type 5, status -5, desc ffff8806c6050a00
> LustreError: 4044:0:(events.c:452:server_bulk_callback()) event type 5, status -5, desc ffff8806b1044800
> LustreError: 4044:0:(events.c:452:server_bulk_callback()) event type 5, status -5, desc ffff8806ae9cd600
> LustreError: 4044:0:(events.c:452:server_bulk_callback()) event type 5, status -5, desc ffff8804d5011000
> LustreError: 4044:0:(events.c:452:server_bulk_callback()) event type 5, status -5, desc ffff880684182800
> LustreError: 4044:0:(events.c:452:server_bulk_callback()) event type 5, status -5, desc ffff8806b103ee00
> 
> _______________________________________________
> lustre-discuss mailing list
> lustre-discuss at lists.lustre.org
> http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org

Cheers, Andreas
--
Andreas Dilger
Lustre Principal Architect
Intel Corporation