Attending: Duncan, Jens (chair+mins), Wenlong, RobC, Teng, Ste, Winnie, RobA, Daniel, Sam

0. No operational issues

   Blog posts - RobC has produced a very informative blog post on CentOS8, so ticking two
   boxes, covering both the CentOS upgrades and  blog posts.

   Tier 1 upgrading CEPH to Nautilus, aiming to finish Q1.  Also of interest to T2s, at
   least - as we have discussed - large ones.

   Also; CEPHfs is now considered production ready, so it may become an option to just
   run CEPH with FS, and a StoRM on top, just like the Lustre sites do.  Or plain xroot
   server.

   The DPM roadmap - contact Sam or myself for the secret link - is being circulated to
   a limited audience (GDB) but should be a public document.  Because the link is a Google
   doc in suggest mode, we are not making it fully public.

   Ditto for the DOMA Access document; Teng shared a link with a forward look towards
   High Luminosity LCG.

   All of these are (perhaps primarily) of interest for T2s to help them decide their future
   paths - should they continue as storage sites, become cache sites, or something else?
   There are additional questions of retaining the durability if we - well, ATLAS - get
   fewer sites storing their data, should the site durability increase - or should a
   decrease be allowed by having better failover like CMS do.  A decrease in durability
   at a single site would be desirable in the sense that it increases the available space
   at the site without impacting the overall durability too much, because there will still
   be enough copies at other sites.

$. NOB


Thanks to a pasteo I lost the chat log, but I still have the links to the DPM whitepaper
and Teng's doc, which were the main pieces of content.  However, as we're not making them
public, contact me if you want the links.