On Wed, Jan 7, 2009 at 01:04, Daryl C. W. O'Shea <[email protected]> wrote: > On 06/01/2009 5:14 AM, Justin Mason wrote: >> On Tue, Jan 6, 2009 at 01:11, Daryl C. W. O'Shea >> <[email protected]> wrote: >>> On 31/12/2008 6:52 AM, Justin Mason wrote: >>>> WDYT? >>> My initial concern is about the semi-occasional breakage the channel has >>> been reportedly experiencing. Is that resolved? >> >> There hasn't been any real breakage > > I was referring, I think mainly, to the occasional reports of access > denied errors when sa-update attempts to download the updates.
What that was, is NFS stale filehandles caused by broken linux NFS :( Basically, the Linux server for yerp.org has multiple separate Xen VMs -- one for SSH connections and one for the httpd. They share the www directory via NFS. When the sought rules are updated, they're written on the SSH-connections VM, but when clients attempt to access them, they're accessed from the httpd VM. Due to thoroughly crappy Linux NFS brokenness, this seems to occasionally get "Stale NFS filehandle" errors. I worked around it over xmas to use rsync to copy over the new files and write them on the httpd VM. this should have sorted that out. If we were to move it off the yerp.org server, this would be avoided anyway. >> -- instead, what's happening is >> that I haven't been doing the required corpus maintainance/rsyncs >> needed to make new ruleset generation necessary, so nothing gets >> published. As far as I can tell.... > > Not quite sure what you mean there. There are occasional mails about "no ruleset has been generated in 8 hours" etc. That's why that's happening. (I think) --j.
