Hi, all:
The login server Willow seems to be having load issues. Several commands are failing because of load issues, it appears. For example, a 'dir' command will return '-bash: fork: Not enough space' about 50% of the time. These issues have been reported by numerous people in the IRC channel over the last 10 or so minutes (as well as tsnag, who was practically spamming).
This is a heads up email, hopefully the ops can take a look at this issue and correct it.
Thank you,
Matthew Bowker matthewrbowker@toolserver.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Yes, e.g. SGE job are not even sent to queue, they go into 'qw' directly after starting... So my bot is down...
Greetings DrTrigon
On 03.03.2012 05:17, Matthew Bowker wrote:
Hi, all:
The login server Willow seems to be having load issues. Several commands are failing because of load issues, it appears. For example, a 'dir' command will return '-bash: fork: Not enough space' about 50% of the time. These issues have been reported by numerous people in the IRC channel over the last 10 or so minutes (as well as tsnag, who was practically spamming).
This is a heads up email, hopefully the ops can take a look at this issue and correct it.
Thank you,
Matthew Bowker matthewrbowker@toolserver.org
_______________________________________________ Toolserver-l mailing list (Toolserver-l@lists.wikimedia.org) https://lists.wikimedia.org/mailman/listinfo/toolserver-l Posting guidelines for this list: https://wiki.toolserver.org/view/Mailing_list_etiquette
On 03/03/12 11:11, Dr. Trigon wrote:
Yes, e.g. SGE job are not even sent to queue, they go into 'qw' directly after starting... So my bot is down...
Greetings DrTrigon
That's the right thing to do. If the server is so overloaded it can't even fork, it makes no sense for SGE to create more processes there. They should be scheduled in some other host.
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
That's the right thing to do. If the server is so overloaded it can't even fork, it makes no sense for SGE to create more processes there. They should be scheduled in some other host.
I thought SGE does itself load balance it and use another host... Anyway it works again now - thanks a lot for this!!
Greetings
toolserver-l@lists.wikimedia.org