The 2014 migration

From DISI
Revision as of 14:20, 13 May 2014 by Frodo (talk | contribs) (Created page with "If you are a member of the lab, or you use our cluster via ssh, big changes are afoot in 2014. This page describes what is happening, a tentative timeline, and answers some qu...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

If you are a member of the lab, or you use our cluster via ssh, big changes are afoot in 2014. This page describes what is happening, a tentative timeline, and answers some questions you may have.

Why

  • Cluster 0 has accumulated structural problems that are not easy to fix in place.
  • Cluster 1 may not be available much longer.
  • Backups have become unwieldy.
  • Cluster maintenance has become very time consuming, and there has been considerable O/S and software drift, leading to complex idiosyncrasies.

What

  • People will move their work from Cluster 0 and Cluster 1 to Cluster 2.
  • Cluster 0 will remain available through the end of 2014.
  • As soon as Cluster 2 can replace 90% of cluster 0, hardware will be repurposed on Cluster 2.

Notes

  • The clusters are being kept completely separate deliberately.
  • Cluster 2 has a higher level of security.

FAQ

  • Do I have to move to Cluster 2 now?

No, right now, we are still recruiting volunteers. When we get to the Cluster 2 = 90% (Cluster 0), we will start moving hardware over to the new cluster. What this means is that Cluster 2 will get more powerful and Cluster 0 less so. This is how we plan to gently incentivize you to move over.

  • What the current CPU counts?

Please see the Category:Cluster page for a summary.