Dream job ad appears: Data wrangler for Square Kilometre Array
Curtin University seeks extreme sysadmin for Murchison 'scope
Australia's Square Kilometre Array boffins – courtesy of Curtin University – have put up the spec for might just be a dream sysadmin job, so long as they're ready, willing and able to cope with what will ultimately be one of the world's biggest data firehoses.
As already recorded by The Register, the Square Kilometre Array precursor, the Murchison Widefield Array, began streaming data last year, and what a stream it is. Back then, the hundreds of tiles totalling 2,048 individual antennas were shipping 400 MB per second back to Pawsey.
The link has since been upgraded to 8 Tbps.
That has to be filtered before it's stored: the raw data of 1.5 TB per hour is thinned down to a mere 3 PB a year (deciding what to throw away is one of the great computing challenges of the SKA project).
Hence Curtin University is looking for someone to work with Pawsey staff to handle that data – an archive that will reach around 9 PB by next year.
Pawsey is soon to be home to a 35,000-core Intel Xeon E5-2600 v3-based machine, the most powerful south of the equator.
The job ad says that the core responsibility will be “the operation and maintenance of the data flow and data archiving between the MWA project and the Pawsey supercomputing centre. Working closely with staff at the Pawsey supercomputing centre you will ensure best practice treatment of the MWA Big Data challenge, with the future view of making MWA data widely and easily available to an international user community.”
Unsurprisingly, a “demonstrated ability to manage the operation and maintenance of a complex large-scale data system is essential”. ®
Sponsored: Hyper-scale data management