[Tetralith-users] Tetralith just got bigger :)

[Mats Kronberg]

Dear Tetralith users,

As you might know, NSC and our hardware vendor have been busy installing
and testing the second phase of Tetralith - an additional 1256 compute
nodes, bringing the total to 1908 nodes.

This work is now done!

We started releasing the new nodes for normal use yesterday, and will
continue to do so, a couple of hundred nodes at a time, over the coming
days (this is done slowly while monitoring the job scheduler and storage
system to make sure we can handle three times the load). At this time,
600 of the 1256 new nodes are available for user jobs.

You do not need to do anything different to use the new nodes, just
submit more jobs. :)

Projects that started January 1st or later already got their correct
computing time allocation (core hours) to match the full size of
Tetralith.

Projects that started before January 1st have been adjusted as follows:

- SNAC Large projects that started before January 1st have had their
  computing time allocation multiplied by 2.9.

- SNAC Medium projects that started before January 1st have had their
  computing time allocation multiplied by 1.5 up to the Medium project
  limit of 200000 core hours/month.

- SNAC Small projects are unchanged.


--
Mats Kronberg
National Supercomputer Centre (NSC)