On Tue, Oct 29, 2024 at 12:42 AM Alfred M. Szmidt <a...@gnu.org> wrote: > > I think this is mixing up things, I was talking about the outages of > overall infrastructure, not Savane. The first is something we have > little control over and is only something the FSF can do something > about. > > There have been several large and long outages the last few months. >
I'm not aware of significant problems since Bob remediated problems blocking Trisquel updates on the git server. This happened around the same time FSF addressed an issue related to power draw/budget. Have there been problems we might not be aware of? (Anything in the last couple of weeks since this work was done?) In any case, we have been working since the move-out from the Boston offices to stabilize things as best we can. IMO the primary constraints are: 1. we are limited to somewhat dated hardware able to run on entirely free software/firmware 2. there are two FTE within the tech team who must resolve all significant infrastructure issues (generally, anything beyond problems within a single VM) 3. nearly all hosts are receiving constant DDoS like activity; a fair amount of staff and volunteer time goes into fighting this off to reserve the compute for valid requests. (See article from Michael in the upcoming issue of the bulletin.) It might be helpful if you document the problems you are thinking of more specifically, given you aren't referring to the outages from before our most recent set of changes (power and VCS server upgrade).
