[IITAC-users] Update on I/O problems

[Paddy Doyle]

Dear all,

The main TCHPC clusters are now fully operational again.

As previously reported, there was a power outage in the machine room on Tuesday
evening of this week. This was due to a UPS failure, which tripped the main
power feed to the room. The TCHPC team was working yesterday and this morning to
implement an alternative for the affected systems.

The work undertaken last week to address the I/O instabilities was a success. We
were able to make the system stable by applying updated device drivers and
system stack on the I/O nodes.

 

The current state of the TCHPC clusters are as follows:

    * Lonsdale: fully operational
    * Parsons: fully operational
    * IITAC: off-line, with some further work required
    * Kelvin: undergoing user testing.


Kind regards,
Paddy

-- 
Paddy Doyle
Trinity Centre for High Performance Computing,
Lloyd Building, Trinity College Dublin, Dublin 2, Ireland.
http://www.tchpc.tcd.ie/