I have set up a git bisect run on mm-crawler01 in hope to find the commit which fixes curl. This means that mm-crawler01 will scan our mirrors in the next few hours continuously. I just want to mention it here in case one of the mirror admins is asking why their mirror is scanned over and over again. It might also put a high load on the database as the crawler has a DB connection open for each thread.
Adrian
infrastructure@lists.fedoraproject.org