ALL NPPS Members

[Maxim Potekhin <potekhin AT bnl.gov>]

FYI, just in case you are not on one of those lists.

-------- Forwarded Message -------- Subject: 3rd + final batch delayed for alma9 shared pool upgrade Date: Fri, 15 Nov 2024 16:05:21 +0000 From: Cowan, Matt <cowan AT bnl.gov> To: liaisons AT rcf.rhic.bnl.gov <liaisons AT rcf.rhic.bnl.gov> CC: linux AT rcf.rhic.bnl.gov <linux AT rcf.rhic.bnl.gov>, RCF Staff <rcfstaff AT rcf.rhic.bnl.gov>

Hi Liasons, we've decided to delay the 3rd and final batch of alma9

shared pool node upgrades a bit.  Originally scheduled to start draining this afternoon (Fri Nov 15, 3pm), we'll delay that until atleast 2pm next Tuesday Nov 19, shortly after the next weekly tuesday 1pm EST alma9 spool upgrade status meeting (zoom link at end).  Depending on feedback received by then, we'll decide if we're able to proceed or further delay.  Ideally we'd still like to finish before Thanksgiving, but we've made good progress and have some wiggle room.
If you have any concerns or alma9 readiness issues, we'd greatly appreciate your attendance at that meeting if possible, and hearing any concerns at that time, or before.
We'd greatly appreciate feedback to the following questions:
  - How would you rate your experiment's alma9 testing/readiness     status/comprehensiveness?  (estimate [0-100]% functionaly ready)     If not yet ready, list of remaining steps or blocking issues appreciated.
  - How would you rate the level of alma9 information disemination     within your experiment?  (estimate [0-100]% users informed)     If your users are not yet fully informed, how can we best coordinate     messaging to avoid confusion and make the transition most seemless?
  - When would be best for your experiment for sdcc to do mass     notifications directing users to new alma9 submit nodes?     email, motd messages on sl7 submit nodes, any other suggested     communication channels?
Reminder: the new alma9 submit nodes are up, running, and usable now, for submitting jobs to alma9 workers.  Existing sl7 submit nodes, and around ~50kHS of to-be-retired sl7 workers, will remain operational until either their retirement, currently targeted for ~2nd week of January, or potential early retirement if all experiments + users have successfully migrated to alma9 earlier?  We've tried to make the alma9 submit node naming simpler and more intuitive; please see the attachment for mapping from old sl7 names to new alma9 names. Note phnxsub01 is a conflict/special case; to avoid confusion it's alma9 version is not yet deployed.
Thanks! -matt

Weekly alma9 shared pool upgrade meeting:  Tuesday, 1pm EST
https://bnl.zoomgov.com/j/1601120984?pwd=zIATubX64kDXv0tq2NRIOmcBx1MNeQ.1

Attachment: Screenshot 2024-11-15 110121.png
Description: PNG image

Attachment: Screenshot 2024-11-13 131309.png
Description: PNG image