On Sat, 2009-06-13 at 14:55 +0930, Tim wrote: > On Fri, 2009-06-12 at 22:20 +0300, Jonathan Dieter wrote: > > The Fedora infrastructure team is trying to streamline the process a > > bit, but the fact remains that generating deltarpms costs a lot in CPU > > time and RAM usage, and the more deltarpms you generate, the more time > > it takes. > > Isn't that done once on a master server, then mirrored? Rather than > each mirror generating them?
Yes, but my point is that the compose time, the time it takes to put together a set of updates, for all of the architectures * (all of the releases + rawhide) has gone from I think three or four hours to 8+ since we started generating deltarpms. If we start generating more deltarpms (which is what I think the OP was asking for), it will go up even more. I doubt the infrastructure team is going to accommodate requests for more deltarpms until the compose time has been brought back down again. Please note that I *do not* speak for the infrastructure team. This is just what I'm picking up from those who are on the infrastructure team. Jonathan
signature.asc
Description: This is a digitally signed message part
-- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
