Print

Print


Hello!
Just a light review this week, I'm just back from leave and have even 
less idea of what's going on then usual! It seems I very quickly forget 
how to type.

The other VO nagios looks a bit of a mess today, but that's being 
handled. We'll look at it next week.

21 Open Tickets this week.

TIER 1
https://ggus.eu/?mode=ticket_info&ticket_id=113914 (26/5)
Sno+ had problems at the Tier 1 where jobs failed whilst uploading data, 
believed to be due to an incorrect VOInfoPath. There's been a failure at 
replicating the issue, and the VOInfoPath advertised is correct. Very 
confusing, as I assume it all worked at some point before! In progress (2/6)

https://ggus.eu/?mode=ticket_info&ticket_id=113910 (26/5)
Another Sno+ ticket, concerning lcg-cp timeouts whilst data-staging from 
tape. Matt M has asked for advice on the best practice for doing this, 
or if Sno+ would be better off just upping their timeouts. Brian has 
given some advice on using the "bringonline" command, but is himself 
unsure the best way of seeing what files are currently in a VO's cache. 
Not much news since. In progress (28/5)

https://ggus.eu/?mode=ticket_info&ticket_id=114004 (31/5)
Atlas transfers fail due to the "bring-online" timeout being exceeded. 
Brian spotted a problem with file timestamps mismatching, but no news on 
this ticket since. In progress (1/6)

BRUNEL
https://ggus.eu/?mode=ticket_info&ticket_id=114006 (31/5)
APEL accounting oddness at Brunel, noticed by the APEL team. After much 
to-and-fro-ing John noticed that multiple CEs were using the same 
SumbitHost, and thus overwriting each other's sync records. Something to 
watch out for. In progress (7/6)

IMPERIAL
https://ggus.eu/?mode=ticket_info&ticket_id=114157 (8/6)
There's been some debate on the atlas lists about this ticket, a classic 
"not enough space at the site" ticket. Raising above the indignation 
over being ticketed for this, Daniela has offered a couple of TB to give 
some space, and pointed out that IC have some atlas data outside space 
tokens, and that this could be used to expand the tokens if cleaned up. 
Waiting for reply (8/6)

And that's all folks!
Cheers,
Matt