2004/07/15
IRC [00:19] *** cd34 joined the chat.
IRC [00:48] <frodoroot> hello?
IRC [03:33] *** frankie joined the chat.
AIM [04:48] *** POLearyUK joined the chat.
AIM [05:14] <POLearyUK> *nobody here but us chickens*
IRC [06:39] <Dossy> bock bock, bock
AIM [06:39] <POLearyUK> dude, it's like 6:45am- go back to bed!
IRC [08:13] *** frankie parted the chat.
IRC [09:35] * leff is back (gone 15:19:14)
IRC [09:37] *** pascals joined the chat.
IRC [09:38] <pascals> Is there a binary distribution for windows somewhere?
AIM [09:40] <POLearyUK> yeah- have a version from someones site, trying to find link to it
AIM [09:40] <POLearyUK> if not i'll put mine up somewhere
IRC [09:40] <pascals> Okay, cool.
IRC [09:41] <pascals> I only have a crippled environment here
IRC [09:41] <pascals> Google doesn't know where it is (or I don't know how to ask), some pointers to the old arsdigita site.
IRC [09:46] *** pascals parted the chat.
AIM [09:46] *** POLearyUK joined the chat.
IRC [09:46] *** rcrit joined the chat.
IRC [09:47] <rcrit> dossy: you up?
IRC [10:27] <Dossy> almost. what's up
IRC [10:28] <Dossy> rcrit - i'll be driving on the road. call my cell ...
IRC [10:29] * Dossy && driving.
AIM [10:43] *** RSeeger00 joined the chat.
IRC [11:38] <Dossy> burp.
AIM [11:40] <RSeeger00> hey
IRC [11:57] *** jader joined the chat.
AIM [12:33] *** POLearyUK joined the chat.
AIM [12:33] <POLearyUK> has anyone had problems with ns_eval sourcing a file only into it's own thread?
IRC [12:37] <jcollins> no
AIM [12:39] <POLearyUK> hmm, wonder has anything changed in ns_eval or are there limits on the ns_tclUnknow lookup table
IRC [12:43] <jcollins> are we talking 4.x?
AIM [12:43] <POLearyUK> yep
IRC [12:43] <jcollins> hrm i don't know then ...
AIM [12:44] <POLearyUK> it's a bit weird- am inclinded to thing we've reached limit in size of proc names or something silly
IRC [12:47] <Dossy> doubtful
IRC [12:47] <Dossy> what's the symtom?
AIM [12:47] <POLearyUK> file can be sourced from cp, and you can do an info body etc there
AIM [12:47] <POLearyUK> but no other thread picks it up
IRC [12:48] <jcollins> and you're ns_evaling it?
AIM [12:49] *** GizmoBeastLives joined the chat.
AIM [12:49] <POLearyUK> yep ns_eval source- even have a web interface for my dev peeps
AIM [12:50] <GizmoBeastLives> ns_eval has changed lots in AOLserver 4.x - have you tried ns_eval -sync?
AIM [12:50] <GizmoBeastLives> http://www.panoptic.com/wiki/aolserver/1306
IRC [12:51] <jcollins> hrm neat
AIM [12:52] <POLearyUK> nice!!! thanks gizmo
AIM [12:52] <GizmoBeastLives> We saw the same thing where the changes on a reinit wouldn't be visible in other threads for maybe 1 minute
AIM [12:52] <GizmoBeastLives> With the -sync, the reinit takes maybe 30 sec, but when it is done, you know everything has updated
AIM [12:53] <POLearyUK> hmm ok- makes a lot of sense
AIM [12:55] *** DossyNJ joined the chat.
IRC [13:02] <Dossy> time to boot VMware to try and fix another AOLsrever bug ...
AIM [13:03] <POLearyUK> VMware -boot mac?
IRC [13:03] <Dossy> no, debian linux
IRC [13:09] <jcollins> does tcl8.5 work fine with aolserver?
IRC [13:09] <Dossy> jcollins - not sure, haven't tried yet. I think RSeeger00 has -- rob?
AIM [13:10] <RSeeger00> Yeah, it works for me
AIM [13:10] *** POLearyUK joined the chat.
AIM [13:10] <RSeeger00> But I haven't really run any exhaustive tests... I just know it runs and it serves pages
IRC [13:10] <jcollins> cool
IRC [13:11] <Dossy> happy: berkeley db is fast. :)
IRC [13:11] *** bartt joined the chat.
IRC [13:11] <Dossy> hi bartt.
AIM [13:13] <RSeeger00> So, 8.5p2
IRC [13:16] <bartt> Hi Dossy
IRC [13:16] <bartt> I've got a question for you
IRC [13:16] <bartt> Are scheduled threads assigned a new number for each time a thread is executed?
IRC [13:19] <bartt> E.g. will a re-occuring thread be reported to the log with a different number each time around?
AIM [13:21] <RSeeger00> If you're talking about threads created with "ns_thread begin..." then I believe the answer is yes, a new thread is created for each one, and with a new number
IRC [13:24] <Dossy> that's not what he's talking about
IRC [13:24] <Dossy> you're talking about ns_schedule_*
IRC [13:24] <Dossy> no, the ID is fixed, AFAIK
IRC [13:26] <jcollins> the job id is fixed
IRC [13:26] <jcollins> the thread id might be different though
IRC [13:26] <jcollins> is my *guess8
IRC [13:28] <jcollins> does the log have the thread id or something else?
IRC [13:29] <Dossy> yeah, if you schedule it with -thread, each time it runs it'll create a new thread -> new thread ID
IRC [13:29] <Dossy> apt-get dist-upgrade taking a long time to download. ugh
IRC [13:30] <jcollins> yikes dist-upgrade
IRC [13:33] <rcrit> is there a htaccess module for aolserver?
IRC [13:36] <Dossy> rcrit: nope.
IRC [13:36] <rcrit> didn't think so.
IRC [13:36] <Dossy> Scary thought. :)
IRC [13:36] <rcrit> ok. the php programs I'm looking at use htaccess for security
IRC [13:36] <Dossy> yeah.
IRC [13:36] <rcrit> htaccess is relatively "easy" to do, if I get time I may write one
IRC [13:36] <Dossy> in the new nsauth module, it might be fun to provide an htaccess-like capability.
IRC [13:37] <Dossy> i really want to trash nsperm and replace with a much more useful nsauth module. blue sky ...
IRC [13:37] <rcrit> yes, nsperm does not seem very good.
IRC [13:37] <Dossy> and the name makes me giggle. *immature*
IRC [13:38] <bartt> On the other hand each conn number in the log corresponds to concurrent connection threads, right?
IRC [13:38] <Dossy> bartt: IF I understand what you just said correctly, yes.
IRC [13:39] <Dossy> rcrit - Jim's been sitting on code to make the url-based req pool routing stuff. should we implement it for him instead?
IRC [13:39] <bartt> What I mean is that if I see [-conn:23] that I can conclude that there were 23 connection threads active at that point in time.
IRC [13:39] <Dossy> bartt: or at some point in time.
IRC [13:39] <rcrit> why not just get his code, clean it up and commit it?
IRC [13:39] <Dossy> bartt: and your request just happened to use #23 in the pool
IRC [13:40] <Dossy> rcrit: step #1 is probably hard. :)
IRC [13:40] <bartt> Dossy: Exactly, the pool could have been larger, but the pool size at that point was at least 23
IRC [13:40] <Dossy> bartt: no, that just means there are 23 threads in the pool, and your request happened to be fulfilled by #23
IRC [13:41] <Dossy> that could have been the only one active at the time, with all other N-1 threads idle
IRC [13:41] <Dossy> two adjacent log entries in my server log:
IRC [13:41] <Dossy> [15/Jul/2004:13:23:19][18046.1095998384][-conn:panoptic.com::423]
IRC [13:41] <Dossy> [15/Jul/2004:13:36:29][18046.1095998384][-conn:panoptic.com::425]
IRC [13:42] <Dossy> oh that's interesting. i bet i've got a bug in my code somewhere - the number keeps going up, and it's hardly doing 1 conn/sec
IRC [13:43] <bartt> I'm inspecting the error logs of a system (that I don't have access to) that terrible performance at some point in time.
IRC [13:44] <Dossy> lovely.
IRC [13:44] <Dossy> your best bet is nstelemetry.adp, *while* the performance suckage is happening.
IRC [13:44] <bartt> Reportedly the number of nsd processes would increase during the performance drop.
IRC [13:44] <bartt> Dossy: I know, but it is too late for that now.
IRC [13:45] <bartt> I will try and recreate it in the office but that will take some time before I can do that.
IRC [13:45] <bartt> So I'm looking at connection numbers and scheduled thread numbers.
AIM [13:47] <DossyNJ> yeah ... going to be tough
IRC [13:48] * bartt sighs
IRC [13:50] <Dossy> yay, spam filter doing kick-ass today.
AIM [14:00] <RSeeger00> Darn, guess that means I have to switch to the ping based DoS attack on your... err, um... nm ;)
IRC [14:09] <frodoroot> a large percentage of the time when I see nsperm I read it n-sperm
AIM [14:09] <GizmoBeastLives> Dossy, the google ads at the bottom of every Wiki page are intentional or work of an unfriendly script?
AIM [14:10] <GizmoBeastLives> I'd hate to think how you read nsexpat :-)
AIM [14:10] <DossyNJ> intentional :)
IRC [14:10] <frodoroot> lo;l
IRC [14:11] *** TheGorilla joined the chat.
IRC [14:13] <frodoroot> hola hermano
IRC [14:13] <TheGorilla> como esta hermanito?
IRC [14:15] <rcrit> Dossy: the minimum stacksize for a "big" php app to work with AOLserver seems to be 256kb, not 1MB.
IRC [14:15] <rcrit> I was going to update the Wiki and say that this is the minimum but we recommend 1MB.
IRC [14:15] <rcrit> and also update the docs with my findings.
IRC [14:16] <frodoroot> TheGorilla: voz esta el hermanito
IRC [14:16] <frodoroot> estas
IRC [14:18] <frodoroot> eres
AIM [14:19] <DossyNJ> rcrit: cool.
AIM [14:19] <DossyNJ> frodoroot: TheGorilla is your big brother? huh?
IRC [14:20] <frodoroot> nah he's the little one
IRC [14:22] <jhavard> unless I'm blind, I wish vlad's download page or something included with nssock and nstcp would state, "This requires my patches to aolserver"
AIM [14:26] <DossyNJ> haha
AIM [14:26] <DossyNJ> :P
IRC [14:26] <jhavard> I mean it says where the patches and pre-patched 4.1 dist are, "This is required for some stuff." but no references to it with nssock and nstcp
IRC [14:33] <rcrit> Dossy: I want to put the perl script that I wrote to use the PHP test cases against AOLserver, how would you recommend I integrate this into the Wiki?
AIM [14:47] <DossyNJ> Hmmm.
AIM [14:47] <DossyNJ> Is the script long?
IRC [14:52] <bartt> If maxconnections is the default of 100, connsperthread 0 and maxthreads 20 how many nsd processes could one possibly see?
IRC [14:52] <rcrit> 371 lines, just over 1k
AIM [15:07] <DossyNJ> rcrit: maybe push it up to www.aolserver.com/ somewhere and put a link to it from the wiki?
AIM [15:07] <DossyNJ> bartt: 20?
IRC [15:10] <bartt> That is what I thought too Dossy.
IRC [15:11] *** cd34 parted the chat.
AIM [17:32] <DossyNJ> okay, i'm heading out for the day
AIM [17:32] <DossyNJ> cheers, everyone
AIM [17:32] *** DossyNJ joined the chat.
AIM [18:39] *** RSeeger00 joined the chat.
AIM [20:17] *** GizmoBeastLives joined the chat.
AIM [23:48] *** DossyShiobara joined the chat.
AIM [23:49] *** DossyShiobara joined the chat.
AIM [23:50] *** DossyShiobara joined the chat.
AIM [23:51] *** DossyShiobara joined the chat.
AIM [23:53] *** DossyShiobara joined the chat.
AIM [23:53] *** DossyShiobara joined the chat.