#musicbrainz-devel

/

      • ijabz joined the channel
      • 2009-09-13 25651, 2009

      • pronik` joined the channel
      • 2009-09-13 25656, 2009

      • brianfreud
        http://codereview.musicbrainz.org/r/391/ ping? 8 days, no review... :)
      • 2009-09-13 25639, 2009

      • luks joined the channel
      • 2009-09-13 25626, 2009

      • MightyJay joined the channel
      • 2009-09-13 25645, 2009

      • outsidecontext joined the channel
      • 2009-09-13 25653, 2009

      • aCiD2 joined the channel
      • 2009-09-13 25627, 2009

      • brianfreud
        aCiD2: trying to follow your last note in CR - are we good then? :D
      • 2009-09-13 25656, 2009

      • brianfreud
        luks: thanks; I just realized that somehow I've not actually been seeing all tickets in CR, as I wasn't part of the ngs group.
      • 2009-09-13 25614, 2009

      • aCiD2
        "good" might be pushing it, but I'll give you approval to carry on :P
      • 2009-09-13 25640, 2009

      • aCiD2
      • 2009-09-13 25617, 2009

      • aCiD2
        seem to be a lot of people I see in perl irc channels advocated its use if you ever want to do caching
      • 2009-09-13 25633, 2009

      • aCiD2
        mind you, they advocate DBIx::Class everywhere, so no guarantee it's useful for MusicBrainz
      • 2009-09-13 25623, 2009

      • luks
        I haven't seen it
      • 2009-09-13 25632, 2009

      • luks
        how would it help us?
      • 2009-09-13 25629, 2009

      • luks
        I generally don't see a reason to change something that works
      • 2009-09-13 25638, 2009

      • aCiD2
        Well we use a memory cache and a memcached cache
      • 2009-09-13 25646, 2009

      • aCiD2
        A unified front end might simplify stuff
      • 2009-09-13 25657, 2009

      • aCiD2
        but I haven't used any of the caching stuff we use atm, nor chi, so I can't comment
      • 2009-09-13 25642, 2009

      • luks
        well, the code is already written and works
      • 2009-09-13 25630, 2009

      • luks
        I'm not saying it's a good idea to use something like this
      • 2009-09-13 25655, 2009

      • luks
        but we already have a working solution, and any change can just bring more problems
      • 2009-09-13 25634, 2009

      • luks
        er, s/it's a good idea/it's not a good idea/
      • 2009-09-13 25633, 2009

      • aCiD2
        sure
      • 2009-09-13 25646, 2009

      • aCiD2
        i'm not really suggesting the change now, no doubt I'll end up having a play with CHI to see what all the hype is about
      • 2009-09-13 25650, 2009

      • aCiD2
        just wanted to let you know too
      • 2009-09-13 25623, 2009

      • outsidecontext joined the channel
      • 2009-09-13 25636, 2009

      • brianfreud
        luks: $(element).attr('name') is a core jQuery function; since when would that not work in IE?
      • 2009-09-13 25650, 2009

      • brianfreud
        :P
      • 2009-09-13 25610, 2009

      • luks
        IE doesn't allow you to set name of a cloned input
      • 2009-09-13 25656, 2009

      • luks
        I know that, because I spent days hopelessly looking for a solution a long time ago
      • 2009-09-13 25656, 2009

      • aCiD2
        Which IE?
      • 2009-09-13 25640, 2009

      • luks
        any IE
      • 2009-09-13 25639, 2009

      • aCiD2
        ouch
      • 2009-09-13 25602, 2009

      • aCiD2
        so luks, you think edit work should let you change artist credit too, right?
      • 2009-09-13 25601, 2009

      • luks
        yes
      • 2009-09-13 25649, 2009

      • aCiD2
        oh, I can't do this now, because we don't have an artist credit editor, heh
      • 2009-09-13 25600, 2009

      • aCiD2
        same with edit release group, which has been stalled for about 2 months now
      • 2009-09-13 25604, 2009

      • luks
        ohh, they somehow managed to workaround the name issue in IE
      • 2009-09-13 25611, 2009

      • luks
        with new jquery it works
      • 2009-09-13 25600, 2009

      • luks
        aCiD2: hm, the work patch is good then
      • 2009-09-13 25613, 2009

      • luks
        but please add some TODO comments there
      • 2009-09-13 25604, 2009

      • aCiD2
        good idea, will do
      • 2009-09-13 25644, 2009

      • luks joined the channel
      • 2009-09-13 25641, 2009

      • djce joined the channel
      • 2009-09-13 25635, 2009

      • outsidecontext joined the channel
      • 2009-09-13 25610, 2009

      • djce
        Anyone using the main web site at the moment who wants to be a guinea pig, please msg me. It involves no work on your part other than confirming whether or not the site keeps working for you. Thanks!
      • 2009-09-13 25655, 2009

      • pronik`
        djce: if it's not too late -- take me ;)
      • 2009-09-13 25612, 2009

      • pronik
        current status -- works
      • 2009-09-13 25605, 2009

      • aCiD2 joined the channel
      • 2009-09-13 25601, 2009

      • pronik
        if anyone wants to call "WTF?!" out loud every five seconds, look here mms://live.msmedia.zdf.newmedia.nacamar.net/zdf/p…
      • 2009-09-13 25618, 2009

      • pronik
        this is the german election fight
      • 2009-09-13 25634, 2009

      • aCiD2` joined the channel
      • 2009-09-13 25618, 2009

      • ruaok joined the channel
      • 2009-09-13 25611, 2009

      • djce
        hey ruaok
      • 2009-09-13 25635, 2009

      • ruaok
        hey
      • 2009-09-13 25650, 2009

      • djce
        It /almost/ worked... but not quite.
      • 2009-09-13 25601, 2009

      • ruaok
        oh, I'm too late, eh?
      • 2009-09-13 25604, 2009

      • ruaok
        what didn't work?
      • 2009-09-13 25616, 2009

      • djce
        For some reason, some requests didn't get the real client IP substituted in place.
      • 2009-09-13 25629, 2009

      • djce
        So for example the ratelimit key used was "ws ip=10.1.1.13"
      • 2009-09-13 25640, 2009

      • djce
        i.e. lots of clients sharing the same ratelimit key,
      • 2009-09-13 25643, 2009

      • djce
        which is bad, of course.
      • 2009-09-13 25654, 2009

      • djce
        Not quite sure where it's going wrong.
      • 2009-09-13 25624, 2009

      • djce
        Anyway, disabled for everyone except me and pronik, who volunteered earlier :-)
      • 2009-09-13 25618, 2009

      • ruaok
        hmmm. was this specific to the approach used or would all approaches have the same problem?
      • 2009-09-13 25649, 2009

      • djce
        The approach used was to minimise downtime,
      • 2009-09-13 25656, 2009

      • djce
        small backwards-compatible steps.
      • 2009-09-13 25620, 2009

      • djce
        Another approach would be a bit of downtime, a bit of mb_server rewriting, and a big bang release.
      • 2009-09-13 25632, 2009

      • djce
        simpler, but far more disruptive.
      • 2009-09-13 25657, 2009

      • ruaok
        "mb_server rewriting" ?
      • 2009-09-13 25606, 2009

      • djce
        The problem is any parts of the code which needs to know the client's IP address.
      • 2009-09-13 25614, 2009

      • djce
        Namely: ratelimiting, and cookie checking.
      • 2009-09-13 25636, 2009

      • djce
        Currently it uses the mod_perl API to query remote_addr,
      • 2009-09-13 25650, 2009

      • djce
        but there's already a hack in place (has been for years) to bodge that,
      • 2009-09-13 25659, 2009

      • djce
        because the client IP is 127.0.0.1 (apache2)
      • 2009-09-13 25603, 2009

      • ruaok
        would it be better for us to do that for NGS and not worry about the old mason codebase?
      • 2009-09-13 25625, 2009

      • djce
        It's a prequisite for the load balancer.
      • 2009-09-13 25632, 2009

      • djce
        So yes, if you're prepared to wait.
      • 2009-09-13 25643, 2009

      • djce
        s/yes/it's an option/
      • 2009-09-13 25610, 2009

      • ruaok
        have you gauged how much work it would be?
      • 2009-09-13 25632, 2009

      • djce
        Well I thought I'd done the work already.
      • 2009-09-13 25641, 2009

      • djce
        So I'm confused as to what I've missed.
      • 2009-09-13 25617, 2009

      • ruaok
        maybe we need a more realistic test sandbox?
      • 2009-09-13 25614, 2009

      • djce
        It's some difference between dexter and stimpy.
      • 2009-09-13 25629, 2009

      • djce
        So it looks like a codebase diff should turn up the problem.
      • 2009-09-13 25647, 2009

      • djce
        it works if the load balancer sends the request to dexter, but not if stimpy.
      • 2009-09-13 25634, 2009

      • ruaok
        I'd start by diffing DBDefs.pm on both boxes.
      • 2009-09-13 25642, 2009

      • ruaok
        the other SVN code *should* be the same.
      • 2009-09-13 25615, 2009

      • djce
        Ah, I think I've found it...
      • 2009-09-13 25624, 2009

      • djce
        some nginx config I was playing with which I forgot to remove.
      • 2009-09-13 25614, 2009

      • djce
        right, I'll give it another go.
      • 2009-09-13 25610, 2009

      • djce
        looks good this time.
      • 2009-09-13 25637, 2009

      • ruaok
        \ø/
      • 2009-09-13 25658, 2009

      • ruaok
        well done!
      • 2009-09-13 25606, 2009

      • djce
        Yes, that was just some config I was trying earlier, seeing if I could short-cut one of the steps.
      • 2009-09-13 25610, 2009

      • djce
        Apparently, I can't :-)
      • 2009-09-13 25623, 2009

      • ruaok
        lo
      • 2009-09-13 25636, 2009

      • djce
        ok, going offline for a moment, back in a couple of minutes
      • 2009-09-13 25645, 2009

      • ruaok
        k
      • 2009-09-13 25645, 2009

      • djce joined the channel
      • 2009-09-13 25652, 2009

      • djce
        ok I'm having some dinner now,
      • 2009-09-13 25659, 2009

      • djce
        ping me if anyone sees any problems
      • 2009-09-13 25614, 2009

      • djce
        and if all's well I'll continue the transition after dinner.
      • 2009-09-13 25624, 2009

      • ruaok
        sounds good.
      • 2009-09-13 25626, 2009

      • ruaok
        happy noms
      • 2009-09-13 25621, 2009

      • navap joined the channel
      • 2009-09-13 25611, 2009

      • ruaok builds updated java lucene indexes
      • 2009-09-13 25620, 2009

      • djce proxies more IPs
      • 2009-09-13 25634, 2009

      • pronik
        damn libunac!
      • 2009-09-13 25601, 2009

      • pronik
        crashes again and again on me and I can't find the workaround luks gave me sometime ago
      • 2009-09-13 25605, 2009

      • djce
        ruaok: everyone's proxying now.
      • 2009-09-13 25610, 2009

      • djce
        seems to be happy.
      • 2009-09-13 25648, 2009

      • ruaok
        sweet!
      • 2009-09-13 25659, 2009

      • ruaok
        I'm going to put the java search server into rotation later today.
      • 2009-09-13 25606, 2009

      • ruaok
        and probably keep if there for a bit.
      • 2009-09-13 25618, 2009

      • ruaok
        the log analysis gave us one bug to take care of.
      • 2009-09-13 25649, 2009

      • djce
        cool, only one. nice
      • 2009-09-13 25619, 2009

      • ruaok
        yeah, paul seemed to think that would work fine.
      • 2009-09-13 25638, 2009

      • ruaok
        the only tricky thing I see is having a recent JVM on all the boxes.
      • 2009-09-13 25647, 2009

      • ruaok
        catbus might be out of date for that.
      • 2009-09-13 25639, 2009

      • djce
        catbus==generator right?
      • 2009-09-13 25653, 2009

      • ruaok
        yep.
      • 2009-09-13 25651, 2009

      • ruaok runs off for a bit to take care of some errands