Resources and Support for Users of the Supercomputing Wales services
The latest Update 4 version of the Intel Parallel Studio XE 2017 has now been installed and made available to all users. This provides Intel Compilers, Intel MPI and Intel MKL as well as other less widely used software. This includes the latest bug fixes and optimizations from Intel, which may be relevant and beneficial […]
In order to replace the firewall at Bangor with a supported & upgraded device, there will be an outage of the Bangor cluster (only) on Friday 7th July from 1200 to 1500. We anticipate it won’t take this long to perform the work, and will return the system to use as soon as possible. Please […]
On Sunday, we suffered an air conditioning failure at Swansea. The compute nodes have therefore been disabled at this time, whilst login and storage remains available. We apologise very sincerely for the inconvenience. Update @1200 Monday: We are awaiting site attendance by the cooling maintenance providers and will update later this afternoon when more news […]
We’re extending the current Swansea cluster outage by a few hours as we spent additional time earlier replacing hardware. We hope to be back by 10pm. Apologies for the inconvenience.
We will have a maintenance outage of the Swansea cluster on Thursday 11th May 2017 from 9am until 5pm. Any jobs running at the start of the outage will be terminated. We will try to retain queued jobs and allow them to launch once service is resumed. Several important pieces of work are scheduled: stage […]
Following yesterday’s planned maintenance, all systems have now returned to normal operation. We also have a new file access capability in beta status. Connecting via SFTP/SCP to scp.hpcwales.co.uk (aka sftp.hpcwales.co.uk) now enables access to all site home and scratch directories direct from outside the HPCW network. The following tables lists the locations of each: $USER […]
There will be a scheduled outage of all HPC Wales Cardiff systems on Tuesday 13th December 2016 from 9am. This will, therefore, cause outage of the Cardiff HTC and Sandy Bridge clusters, plus also the Access Nodes (login.hpcwales.co.uk) and File Transfer service (scp.hpcwales.co.uk / sftp.hpcwales.co.uk). Any jobs running at the start of the outage will […]
The upgraded software stack that has been deployed over the last 8 months to HPCW systems has proven reliable and capable. Stack related issues have reduced substantially and system availability/user experience has improved as a result. It is therefore time to complete the transition to the new stack by upgrading the Cardiff HTC cluster to […]
Home file store is a key part of the HPCW infrastructure. Unfortunately, the service deployed for this purpose has now reached and gone beyond the end of its manufacturer’s service and supported life. As such, in order that we can utilise a supported and maintained service, we must migrate to a new service. We will […]
The Bangor system has been upgraded to the new software stack. This represents a substantial upgrade for the Bangor system and it will extend its useful life significantly. As an HPCW Phase 1 system based on Intel Westmere technology, it was previously running Red Hat/CentOS 5.5 and the IBM Platform LSF cluster scheduler. The new […]