Karsten 'quaid' Wade wrote:
On Fri, Nov 07, 2008 at 11:43:30AM -0500, Michael DeHaan wrote:
The existing scanner code is pretty solid. I'm happy that it's not regex based in that we can be pretty accurate.
However, if we instead process the tarballs, it turns out that the txt.gz file is txt, not HTML, so we run into problems.
I'm really not looking forward to rewriting the scanner to use complicated regexen again.
I would reason that -- if we know not to even /look/ at previous months we've only scanned -- we use DB transactions (so a month of Fedora-devel is only one SQL statement)
then the daily cost of running the scanner even if it does download individual message pages will be acceptably low.
The only pain will be running them back to the dawn of mailing list time.
I'm going to try to the above first and then we can also see if another db is faster, though I bet if we use transactions even sqlite is sufficiently fast. TBD.
Are we ok with the above?
It sounds sane, given that we'll have a dedicated server working on it. We can then plan the direction of the code based on it being an always-running service v. an occasional snap-shotter.
I will note that there are some tools specifically designed for searching mail archives, such as 'grepmail'. That is, since mail is a standard format and we are only looking for things found in the headers, maybe the regex wouldn't need to be present or so evil
Nice... wasn't aware of that one.
.
I would only consider doing that if we wanted a dual-purpose tool, that is good for snapshots-from-scratch as well as ongoing data mining.
Side topic -- Max, if Infra is concerned about staff power, I am a junior member of that team who has never tried getting sponsored for something. I could pitch in to help design, define, provision, and support our system. That would give us the advantage of having someone within the team who can e.g. tweak cron settings, roll out new code to the server, etc. In the case of a publictest server, the developer does have that control, but once it moves to Infra control, things get all Puppety. :)
Would we need RPM packaging and inclusion in Fedora repos for that? I don't believe all of the infrastructure tools are packaged.
Mainly I'm pondering how we update the EKG instance. If it's something we have SSH to, it's just as simple as "git pull", which I like, for something in active development.
If anyone has /any/ spare server around for this purpose, that might be somewhat easier than having to contend with Puppet for something this simple.
Thoughts?
- Karsten
ekg mailing list ekg@lists.fedorahosted.org https://fedorahosted.org/mailman/listinfo/ekg