00:04:49  * piscisaureus_joined
00:30:23  * sh1mmerquit (Quit: sh1mmer)
01:09:13  <mjr_>ryah: the bnoordhuis stats checking patch seems to be working for us
01:09:21  <mjr_>We've been running it for 2 days now, and no lockups.
01:10:12  <piscisaureus_>and what about the leaks?
01:10:29  <mjr_>Oh yeah, I'm tempted to say that it also fixed those.
01:10:40  <mjr_>Just looking at the memory graphs now.
01:23:08  * mikealquit (Quit: Leaving.)
01:24:34  * antislurpjoined
01:25:10  <antislurp>(Log server was down. This is what it missed:)
01:25:12  <antislurp>[02:13] <mjr_> Yeah, memory is looking WAY better
01:25:14  <antislurp>[02:13] <mjr_> https://skitch.com/mranney/gxk9c/voxer-metrics
01:25:15  <antislurp>[02:13] <mjr_> Looks like we also have some bugs which I need to fix.
01:25:17  <antislurp>[02:13] <mjr_> A lot more crashes than I was even aware of.
01:25:18  <antislurp>[02:13] <mjr_> But memory never goes above 290MB, which I'm pretty happy about.
01:25:20  <antislurp>[02:14] <mjr_> So, hooray.
01:25:21  <antislurp>[02:14] <piscisaureus_> good to know
01:25:23  <antislurp>[02:15] <piscisaureus_> apparently the leak was also caused by the status bug then
01:25:24  <antislurp>[02:15] <mjr_> Yeah, I think so.
01:25:26  <antislurp>[02:15] <mjr_> BTW, we have 470K devices connected to our cluster at the moment. Anybody who says that node doesn't scale is doing it wrong.
01:25:27  <antislurp>[02:16] <piscisaureus_> nice :-)
01:25:29  <antislurp>[02:16] <piscisaureus_> well it depends obviously... node doesn't scale, you have to do it yourself
01:25:30  <antislurp>[02:16] <piscisaureus_> how many clients do you have per node instance?
01:25:32  <antislurp>[02:16] <mjr_> Every day we push that number higher. If this keeps up, we'll be doing 1 million devices at a time.
01:25:33  <antislurp>[02:16] <mjr_> More than a thousand, less than 2 thousand.
01:25:35  <antislurp>[02:17] <mjr_> Yeah, but I'd argue that nothing scales by itself, really. You always have to do some work and think carefully about it.
01:25:45  * antislurppart
01:27:14  * mralephquit (Quit: Leaving.)
01:27:44  <piscisaureus_>mjr_: is that 1000 < n < 2000 number a semi arbitrary choice?
01:28:06  <piscisaureus_>mjr_: or can't node handle more?
02:05:59  <mjr_>I thought for sure it could handle more, but the TLS stuff seems to bog it down.
02:06:08  <mjr_>It's kind of hard to tell where the CPU is going though.
02:06:37  <mjr_>But right now, they are running at around 50% CPU on each core at peak, and much more than that and the latency starts to suffer.
02:09:54  * slurp1joined
02:10:02  * slurp1quit (Read error: Connection reset by peer)
02:11:31  * brsonquit (Quit: leaving)
02:14:39  * sh1mmerjoined
02:39:59  <ryah>mjr_: great news
02:40:11  <mjr_>Yeah, it's pretty awesome
02:40:40  <ryah>i'll land that patch and roll out 0.6.6
02:40:51  <mjr_>Cool.
02:41:05  <mjr_>Whenever you want me to test some HTTPS performance improvement fixes, I'll be all over that.
02:42:54  <ryah>i can't wait to start running profiling with dtrace accross your whole cluster
02:43:05  <mjr_>Yeah, it's going to be awesome.
02:43:20  <mjr_>Needs a few more weeks and bunch more migration pain, but we'll get there.
02:43:26  <ryah>we'll see which js or c++ functions are hot
02:43:43  <mjr_>Since our burning issue right now is riak, not node.
02:44:17  <mjr_>BTW, remember how we had to use that awful patch to ignore ETIMEDOUT and some other errors on HTTPS sockets?
02:44:21  <mjr_>0.6 apparently fixes this.
02:44:52  <ryah>that's good
02:46:11  <mjr_>We do get an awful lot of these though: error: connect Unknown system errno 110
02:46:21  <mjr_>Which I guess is ETIMEDOUT
02:46:38  <mjr_>Not on HTTPS endpoints though, so I dunno WTF is going on.
03:22:29  * isaacsquit (Quit: isaacs)
04:13:32  * piscisaureus_quit (Ping timeout: 240 seconds)
04:18:43  * piscisaureus_joined
04:37:01  * sh1mmerquit (Quit: sh1mmer)
04:48:38  * piscisaureus_quit (Quit: ~ Trillian Astra - www.trillian.im ~)
04:51:34  * piscisaureus_joined
05:13:00  * piscisaureus_quit (Ping timeout: 252 seconds)
05:51:50  * isaacsjoined
06:12:57  * isaacsquit (Quit: isaacs)
06:58:01  * ericktquit (Quit: erickt)
07:14:17  * mikealjoined
07:51:17  * mikealquit (Quit: Leaving.)
07:51:54  * mikealjoined
08:43:42  <txdv>hello
08:43:47  <txdv>is there a way to test only one specific test?
09:01:51  <txdv>I changed some function definitions and i am not too fond of changing every single test
09:02:02  <txdv>ffs i just cant find a way to make only one test run
09:03:57  <txdv>I am removing them from the makefile and I end up with messages like this: /tmp/cck4ITDt.o:(.data+0xb8): undefined reference to `run_helper_tcp6_echo_server'
09:04:11  <txdv>and I am like wtf, cause there is no file containing a function call like that
09:06:29  <txdv>very very fustrating :(
09:13:17  * paddybyersquit (Quit: paddybyers)
09:13:50  * paddybyersjoined
09:18:09  <indutny>txdv: try tools/test.py simple/test-eval
09:18:25  <indutny>or are you speaking about libuv?
09:18:54  <txdv>yes i am
09:19:01  <indutny>txdv: ok, one second
09:20:13  <indutny>test/run-tests fs_file_async fs_file_async
09:21:14  <txdv>yeah
09:21:24  <txdv>now i am facing the problem that all but one test doesn't compile
09:21:34  <txdv>so run-tests doesn't compile either
09:21:45  <txdv>but i dont care about the other tests, i just want to get that one little test running
09:25:01  <txdv>only one test compiles*
09:25:30  <txdv>so I removed from the makefile all the other test files
09:25:37  <txdv>in TESTS=test/test-*
09:25:50  <txdv>but the problem is that if i want to compile that runner, it complains about the undefined references
09:26:11  <txdv>and I can't find out where the heck they are defined
09:27:57  <txdv>http://paste.ubuntu.com/766698/
09:31:41  <txdv>ok I found the freaking list definition
09:31:42  <txdv>nevermind
09:53:58  * paddybyersquit (Quit: paddybyers)
10:17:19  <txdv>freaking code
10:17:20  <txdv>:(
10:17:56  <pquerna>would people be okay adding a handle type of something like 'custom', which defined a few new struct members with callbacks for their operations like _close()
10:18:34  <pquerna>then in the switch (handle) in uv__finish_close for example, call said callbacks
10:26:48  * mralephjoined
11:16:00  * piscisaureus_joined
11:16:53  * piscisaureus__joined
11:20:24  * piscisaureus_quit (Ping timeout: 252 seconds)
11:23:03  <txdv>libuv is a bitch to port to .net
11:25:30  <indutny>ooh
11:26:04  <indutny>stupid zerigo dns... I've no words
11:26:13  <txdv>zerigo dns?
11:28:19  <txdv>can't we move the UV_REQ_FIELDS to the end of the definitions, so one could use the same struct definition for linux and windows?
12:08:11  <txdv>libuv devs, y u not develop libuv 24/7!? :D
12:56:18  * mraleph1joined
12:57:33  * mralephquit (Read error: Connection reset by peer)
13:03:15  * mraleph1quit (Ping timeout: 252 seconds)
14:50:24  * paddybyersjoined
14:55:07  * paddybyersquit (Client Quit)
15:22:24  * isaacsjoined
16:02:09  <indutny>ryah: ?
16:02:13  <indutny>good morning!
16:28:35  * paddybyersjoined
16:29:47  * ericktjoined
16:51:42  * mralephjoined
16:54:24  * paddybyersquit (Quit: paddybyers)
17:05:04  * ericktquit (Quit: erickt)
17:52:32  * ericktjoined
17:54:06  * isaacsquit (Quit: isaacs)
18:18:33  * dshaw_joined
18:38:05  * AndreasMadsenjoined
19:06:20  * isaacsjoined
19:13:51  * mjr_quit (Quit: mjr_)
19:14:41  * isaacsquit (Quit: isaacs)
19:14:54  * isaacsjoined
19:22:43  * ericktquit (Quit: erickt)
19:53:26  * mikealquit (Quit: Leaving.)
20:08:40  * mralephquit (Quit: Leaving.)
20:17:09  * AndreasMadsenquit (Remote host closed the connection)
20:48:19  * brsonjoined
21:22:50  * isaacschanged nick to isaacs_away
21:24:55  * piscisaureus__quit (Quit: ~ Trillian Astra - www.trillian.im ~)
21:52:00  * sh1mmerjoined
21:53:35  * mikealjoined
21:53:39  * mikealquit (Client Quit)
21:56:22  * paddybyersjoined
22:04:04  * ljacksonquit (Read error: Operation timed out)
22:18:12  * creationixquit (*.net *.split)
22:18:14  * igorziquit (*.net *.split)
22:18:14  * sh1mmerquit (*.net *.split)
22:18:14  * dshaw_quit (*.net *.split)
22:18:15  * indutnyquit (*.net *.split)
22:18:17  * ircretaryquit (*.net *.split)
22:18:18  * brsonquit (*.net *.split)
22:18:18  * isaacs_awayquit (*.net *.split)
22:18:20  * AvianFluquit (*.net *.split)
22:18:21  * raggi_quit (*.net *.split)
22:18:21  * DrPizzaquit (*.net *.split)
22:18:24  * mmaleckiquit (*.net *.split)
22:18:25  * txdvquit (*.net *.split)
22:18:26  * ryahquit (*.net *.split)
22:18:26  * CIA-111quit (*.net *.split)
22:18:28  * pquernaquit (*.net *.split)
22:18:29  * benviequit (*.net *.split)
22:18:29  * einarosquit (*.net *.split)
22:19:05  * indutnyjoined
22:19:06  * sh1mmerjoined
22:19:06  * brsonjoined
22:19:06  * isaacs_awayjoined
22:19:06  * dshaw_joined
22:19:06  * AvianFlujoined
22:19:06  * igorzijoined
22:19:06  * DrPizzajoined
22:19:06  * raggi_joined
22:19:34  * mmaleckijoined
22:20:01  * ircretaryjoined
22:20:01  * creationixjoined
22:20:10  * ryahjoined
22:20:10  * CIA-111joined
22:20:18  * dshaw_quit (Quit: Leaving.)
22:20:49  * pquernajoined
22:21:27  * txdvjoined
22:25:34  * benviejoined
22:25:34  * einarosjoined
22:25:39  * mralephjoined
22:34:12  * dshaw_joined
22:48:27  * dshaw_quit (Quit: Leaving.)
22:53:50  * dshaw_joined
22:55:27  * mralephquit (Quit: Leaving.)
22:57:01  * dshaw_quit (Read error: Connection reset by peer)
22:58:08  * dshaw_joined
23:01:25  * paddybyersquit (Quit: paddybyers)