Displaying #traffic-server/2015-11-03.log:

Tue Nov 3 00:07:15 2015  gancho:Joined the channel
Tue Nov 3 00:12:24 2015  msekimura:Joined the channel
Tue Nov 3 00:32:41 2015  gancho:Joined the channel
Tue Nov 3 01:24:40 2015  _klk_:Joined the channel
Tue Nov 3 01:30:23 2015  _klk_:Joined the channel
Tue Nov 3 01:42:52 2015  gancho:Joined the channel
Tue Nov 3 02:03:11 2015  kichan:briang , we have a patch for TS-3991
Tue Nov 3 02:03:55 2015  kichan:it is sort of hard for us to share the code with you, though
Tue Nov 3 02:04:16 2015  kichan:but the patch is pretty trivial and seems to work for us
Tue Nov 3 02:28:58 2015  danielxu:Joined the channel
Tue Nov 3 02:34:01 2015  danielxu:Joined the channel
Tue Nov 3 02:56:55 2015  briang:kichan, can you attach it to the ticket
Tue Nov 3 03:22:14 2015  gancho:Joined the channel
Tue Nov 3 03:30:51 2015  gancho:Joined the channel
Tue Nov 3 03:44:22 2015  es:Joined the channel
Tue Nov 3 04:14:57 2015  gancho:Joined the channel
Tue Nov 3 04:47:57 2015  zwoop:Anyone recognize this?
Tue Nov 3 04:48:10 2015  zwoop:In file included from ../../../lib/ts/Diags.cc:44:
Tue Nov 3 04:48:11 2015  zwoop:../../../lib/ts/Diags.h:272:10: error: private field 'start_time' is not used [-Werror,-Wunused-private-field]
Tue Nov 3 04:48:11 2015  zwoop: time_t start_time; // time since Diags was constructed
Tue Nov 3 04:48:31 2015  zwoop:bcall ^ ?
Tue Nov 3 05:02:16 2015  zwoop:Joined the channel
Tue Nov 3 05:04:02 2015  jpeach:Joined the channel
Tue Nov 3 05:05:14 2015  danielxu:Joined the channel
Tue Nov 3 05:06:17 2015  zwoop:Joined the channel
Tue Nov 3 05:09:22 2015  zwoop:amc I think that might be yours ?
Tue Nov 3 05:15:48 2015  jpeach:Joined the channel
Tue Nov 3 05:16:37 2015  gancho:Joined the channel
Tue Nov 3 05:22:52 2015  bcall:Joined the channel
Tue Nov 3 05:23:22 2015  briang:Joined the channel
Tue Nov 3 05:23:23 2015  briang:Joined the channel
Tue Nov 3 05:25:17 2015  psp:Joined the channel
Tue Nov 3 05:25:25 2015  dcarlin:Joined the channel
Tue Nov 3 05:32:46 2015  gancho:Joined the channel
Tue Nov 3 05:33:04 2015  psp1:Joined the channel
Tue Nov 3 05:39:33 2015  _klk_:Joined the channel
Tue Nov 3 05:40:26 2015  gancho:Joined the channel
Tue Nov 3 06:36:16 2015  gancho:Joined the channel
Tue Nov 3 06:48:34 2015  kichan:Joined the channel
Tue Nov 3 06:57:08 2015  gancho:Joined the channel
Tue Nov 3 06:59:42 2015  gancho:Joined the channel
Tue Nov 3 07:01:35 2015  gancho:Joined the channel
Tue Nov 3 07:10:00 2015  gancho:Joined the channel
Tue Nov 3 07:11:33 2015  gancho:Joined the channel
Tue Nov 3 07:13:27 2015  gancho:Joined the channel
Tue Nov 3 07:15:19 2015  gancho:Joined the channel
Tue Nov 3 07:29:43 2015  ryan:Joined the channel
Tue Nov 3 07:29:51 2015  ryan:hello
Tue Nov 3 07:30:19 2015  Guest35549:may i ask a question about SSDcache
Tue Nov 3 07:30:44 2015  Guest35549:base on this link: https://cwiki.apache.org/confluence/display/TS/SSDSupport
Tue Nov 3 07:31:16 2015  Guest35549:while installing ATS i must use : sudo ./configure --prefix=/opt/ts --enable-interim-cache
Tue Nov 3 07:31:49 2015  Guest35549:but it shows me the error: configure: WARNING: unrecognized options: --enable-interim-cache
Tue Nov 3 07:32:22 2015  danielxu:Joined the channel
Tue Nov 3 07:33:08 2015  Guest35549:any idea?
Tue Nov 3 07:47:47 2015  Guest35549:any idea?
Tue Nov 3 07:48:08 2015  felicity:probably this was an experimental feature that isn't merged
Tue Nov 3 07:48:20 2015  Guest35549:is it enabled by default in version 6?
Tue Nov 3 07:48:24 2015  felicity:since that's what that section of the wiki is for
Tue Nov 3 07:48:56 2015  felicity:it's not in 6.0 that i know of, i don't remember seeing it on the changelog
Tue Nov 3 07:49:53 2015  Guest35549:so whats the best way to use SSD and ordinary HDD in mixed mode?
Tue Nov 3 07:50:23 2015  felicity:you could use a filesystem that has that support built-in (like ZFS)
Tue Nov 3 07:50:49 2015  felicity:although large TS caches in files (as opposed to raw devices) aren't usually recommended
Tue Nov 3 08:01:14 2015  Guest35549:but setting exist in latest config file: proxy.config.cache.interim.storage
Tue Nov 3 08:02:01 2015  Guest35549:is there any email support from developers to i ask?
Tue Nov 3 09:16:49 2015  Lethalman:Joined the channel
Tue Nov 3 09:37:05 2015  PSUdaemon:amc: he is
Tue Nov 3 09:59:33 2015  danielxu:Joined the channel
Tue Nov 3 10:04:40 2015  danielxu:Joined the channel
Tue Nov 3 11:03:02 2015  ushachar:Joined the channel
Tue Nov 3 11:24:03 2015  JSeymour:Joined the channel
Tue Nov 3 11:24:07 2015  JSeymour:G'day
Tue Nov 3 11:24:56 2015  JSeymour:Joined the channel
Tue Nov 3 11:53:46 2015  Lethalman:Joined the channel
Tue Nov 3 11:55:01 2015  bahumbug:Joined the channel
Tue Nov 3 11:55:01 2015  bahumbug:Joined the channel
Tue Nov 3 12:09:16 2015  JSeymour:Joined the channel
Tue Nov 3 12:18:03 2015  asldjaljd:Joined the channel
Tue Nov 3 12:18:16 2015  asldjaljd:400 Response Not Cachable getting this error when I try to push!
Tue Nov 3 12:18:41 2015  asldjaljd:though I have configured my APS to ignore all the headers and cache everything
Tue Nov 3 13:01:11 2015  dynek:Joined the channel
Tue Nov 3 13:35:11 2015  dynek:Joined the channel
Tue Nov 3 13:58:48 2015  shinrich1:Joined the channel
Tue Nov 3 14:29:49 2015  jrushford:Joined the channel
Tue Nov 3 14:33:31 2015  esproul:Joined the channel
Tue Nov 3 14:40:16 2015  jrushford:Joined the channel
Tue Nov 3 14:44:58 2015  davet_:Joined the channel
Tue Nov 3 14:53:18 2015  [Jok]:Joined the channel
Tue Nov 3 15:02:49 2015  amc:Joined the channel
Tue Nov 3 15:16:27 2015  sudheerv:Joined the channel
Tue Nov 3 15:48:07 2015  jpeach:TS-3541
Tue Nov 3 16:22:34 2015  bahumbug:Joined the channel
Tue Nov 3 16:39:56 2015  es:Joined the channel
Tue Nov 3 17:12:13 2015  es1:Joined the channel
Tue Nov 3 17:17:30 2015  msekimura:Joined the channel
Tue Nov 3 17:25:34 2015  gancho:Joined the channel
Tue Nov 3 17:32:12 2015  blattj1:Joined the channel
Tue Nov 3 17:33:16 2015  Becoming:Joined the channel
Tue Nov 3 17:33:57 2015  blattj:Joined the channel
Tue Nov 3 17:53:59 2015  jrushford:Joined the channel
Tue Nov 3 18:13:20 2015  jpeach:hey jsime?
Tue Nov 3 18:14:09 2015  psp1:Joined the channel
Tue Nov 3 18:14:23 2015  jsime:jpeach: yes?
Tue Nov 3 18:14:34 2015  jpeach:thanks for the docs patch :)
Tue Nov 3 18:14:57 2015  jsime:glad to have that merge behind us now
Tue Nov 3 18:15:04 2015  jpeach:looks like something went wrong on RTD with the theme :(
Tue Nov 3 18:15:14 2015  jsime:hmm, one sec
Tue Nov 3 18:15:24 2015  jpeach:https://docs.trafficserver.apache.org/
Tue Nov 3 18:15:58 2015  jsime:yeah, looks like our overrides.css got in there (to fix the table wrapping issues), but the two RTD CSS files that should be there... aren't
Tue Nov 3 18:16:22 2015  psp2:Joined the channel
Tue Nov 3 18:17:53 2015  jsime:might be i need to add those explicitly to the list at conf.py:236
Tue Nov 3 18:29:46 2015  jpeach:do you have time budget to look into it this week?
Tue Nov 3 18:44:05 2015  jpeach:thanks jsime
Tue Nov 3 18:44:21 2015  jpeach:zwoop: https://docs.trafficserver.apache.org/css/sphinx_rtd_theme.css is a 404 ... should it be?
Tue Nov 3 18:44:47 2015  psp1:Joined the channel
Tue Nov 3 18:48:18 2015  psp1:Joined the channel
Tue Nov 3 18:53:05 2015  jsime:jpeach: both of the the rtd CSS should be hosted on https://media.readthedocs.org if we want to keep current with any of their tweaks - that's where previous versions of the docs were getting them from, at least
Tue Nov 3 18:53:33 2015  jsime:we could pull those and host them from docs.trafficserver.apache.org along with all the other files if we want, i suppose
Tue Nov 3 18:54:56 2015  jpeach:yeh I was curious because zwoop set up docs.trafficserver.a.o to front it, so wondering if that should cache the CSS as well
Tue Nov 3 18:55:35 2015  jpeach:it actually 404s and sends a set-cookie for the wrong domain :( http://fpaste.org/286556/76918144/
Tue Nov 3 18:56:16 2015  jpeach:jsime: after your fix http://trafficserver.readthedocs.org looks good, but docs.trafficserver.a.o is still serving a cached copy of the broken page :(
Tue Nov 3 18:56:58 2015  jpeach:jsime: do you have a pipeline of docs updates from zwoop et al?
Tue Nov 3 19:02:14 2015  jsime:jpeach: there's the open jira issues for right now, but the plan was also to get together with zwoop & friends to go through that list and some other areas of improvement once the big merge was taken care of
Tue Nov 3 19:02:59 2015  jsime:which with that now being the case, i need to ping the 'et al.'s and see about scheduling that
Tue Nov 3 19:07:48 2015  jpeach:now that the first pass has landed I think we need to make a second pass on structure
Tue Nov 3 19:09:19 2015  jsime:sensible - and should be much easier to get good feedback now with one feeble attempt now visible/dissectable
Tue Nov 3 19:09:46 2015  jpeach:I'll try to find some time to write something down
Tue Nov 3 19:30:40 2015  shinrich1:Joined the channel
Tue Nov 3 19:34:55 2015  _klk_:Joined the channel
Tue Nov 3 20:18:12 2015  jrushford:Has something changed with logging recently? I'm working with the master branch and i've noticed that logging seems to be buffered now.
Tue Nov 3 20:19:14 2015  jrushford:I see in traffic.out that a log message will stop in the middle and then I'll see two or three messages and then it stops in the midle of another message.
Tue Nov 3 20:34:28 2015  amc:Yes, we pushed up the log rotation PR.
Tue Nov 3 20:34:28 2015  amc:I will try to take a look at that. It's a non-trivial change so there's still a few kinks in it.
Tue Nov 3 20:36:17 2015  amc:jrushford - does the rest of the message eventually show up? Is this in traffic.out / diags.log or another log file?
Tue Nov 3 20:37:23 2015  jrushford:yes the rest of the message eventually shows up.
Tue Nov 3 20:38:07 2015  jrushford:it behaves as though its buffered and when the buffer fills up, the messages appear to flush
Tue Nov 3 20:38:13 2015  jrushford:does that make sense?
Tue Nov 3 20:40:25 2015  jrushford:so if you set the debug string in records.config to something that won't log a ton of messages, you'll see it.
Tue Nov 3 20:41:17 2015  jrushford:i'm working on my parent selection changes and I have the debug string set to 'parent_select'.
Tue Nov 3 20:42:55 2015  psp:Joined the channel
Tue Nov 3 20:45:50 2015  jrushford:ams - that was in traffic.out btw.
Tue Nov 3 20:48:35 2015  amc:So this is for traffic.out, right?
Tue Nov 3 20:48:50 2015  amc:Yeah, I suspect the code is opening the files in buffered mode, which is the default. I'll take a look.
Tue Nov 3 20:49:57 2015  jrushford:thanks amc!
Tue Nov 3 22:00:22 2015  tomr_frnd:Joined the channel
Tue Nov 3 22:01:51 2015  shinrich2:Joined the channel
Tue Nov 3 22:10:14 2015  shinrich1:Joined the channel
Tue Nov 3 22:17:16 2015  phibs:Joined the channel
Tue Nov 3 22:17:59 2015  dxu:Joined the channel
Tue Nov 3 22:20:22 2015  dxu:Joined the channel
Tue Nov 3 23:13:13 2015  zwoop:amc you looking at the build failures ?
Tue Nov 3 23:14:23 2015  zwoop:amc first these: https://ci.trafficserver.apache.org/files/clang-analyzer/2015-11-03-091755-18575-1/
Tue Nov 3 23:14:35 2015  zwoop:and also a bunch of failures in Diags.h/.cc due to the unused member variable
Tue Nov 3 23:16:45 2015  davet_:Left the channel
Tue Nov 3 23:34:17 2015  amc:Gah. I thought I had the build fixed. Sigh.
Tue Nov 3 23:35:51 2015  amc:I'll work on that.
Tue Nov 3 23:38:27 2015  zwoop:there's also a slew of coverity issues on the same commit
Tue Nov 3 23:38:30 2015  zwoop:7 of them
Tue Nov 3 23:38:43 2015  dxu:For the unusded member variable in Diags, it was accidentally left in there. Originially I intended to use that, but I went a different route. That can be removed without any issues
Tue Nov 3 23:38:44 2015  zwoop:amc I'd suggest either back it out, and try again later, or try to get this fixed asap IMO.
Tue Nov 3 23:39:20 2015  zwoop:amc there's just a lot of breakage from this commit unfortunatelhy :/
Tue Nov 3 23:56:19 2015  zwoop:back
Tue Nov 3 23:56:50 2015  PSUdaemon:seems like it should be easy to fix up
Tue Nov 3 23:56:56 2015  zwoop:dxu yeah, that'd be great if you can just remove it. Also take a look at the coverity issues (there are 7 of them) and the clang-analyzer issues (3 of them)
Tue Nov 3 23:57:03 2015  PSUdaemon:just a little bit sloppy of a commit
Tue Nov 3 23:57:07 2015  PSUdaemon:but not the end of the world
Tue Nov 3 23:57:25 2015  zwoop:yeah
Tue Nov 3 23:58:16 2015  zwoop:Thomas had started looking at being able to run some builds against a github branch, but I'm a little weary on that since it kinda opens up major attack vectors to hack our CI system.

Comments