Displaying #cassandra-dev/2017-05-16.log:

Tue May 16 00:00:42 2017  mebigfatguy:Joined the channel
Tue May 16 00:05:36 2017  kohlisankalp:Joined the channel
Tue May 16 00:16:21 2017  exlt:heh
Tue May 16 00:16:50 2017  exlt:I just did a build in site/src and got nothing different but a few <code class="highlighter-rouge"> lines
Tue May 16 00:16:58 2017  exlt:what's not current rustyrazorblade?
Tue May 16 00:18:26 2017  exlt:shakes fist at jekyll
Tue May 16 00:18:47 2017  exlt:s/div/pre/ in a few places, too..
Tue May 16 00:23:37 2017  dikang:Joined the channel
Tue May 16 00:28:56 2017  kohlisankalp:Joined the channel
Tue May 16 00:44:55 2017  rustyrazorblade:exlt: well, for example, https://cassandra.apache.org/doc/latest/ should have a search box. i added that back in Feb. 9bd482d36b0c48f62120b260d69c7f25d1abf6eb
Tue May 16 00:45:14 2017  rustyrazorblade:afaik the docs haven't been rebuilt since before feb 22
Tue May 16 00:48:44 2017  rustyrazorblade:another example, https://cassandra.apache.org/doc/latest/operating/repair.html
Tue May 16 00:49:09 2017  exlt:I must be doing this wrong, all I've ever done was run `make` to update pages
Tue May 16 00:49:16 2017  rustyrazorblade:Feb 23 a very thorough description of repair was added: 4cfaf855c404256a9dd281d5066cc076232d72ff
Tue May 16 00:49:29 2017  exlt:and jekyll does a lot of formatting changes that have no content changes
Tue May 16 00:49:39 2017  rustyrazorblade:where are you running make?
Tue May 16 00:51:04 2017  exlt:I have the full cassandra svn checked out && cd site/src/ && make
Tue May 16 00:51:33 2017  exlt:it removes things that shouldn't be removed, changes to html tags for highlighting - no content changes
Tue May 16 00:52:01 2017  exlt:then I have to `svn revert --recursive` things and look through the diff for actual changes
Tue May 16 00:52:06 2017  exlt:annoying as fsck
Tue May 16 00:53:09 2017  driftx:let pcmanus check it out tomorrow
Tue May 16 00:53:14 2017  rustyrazorblade:yeah i was going to say
Tue May 16 00:53:27 2017  rustyrazorblade:he set it up to begin
Tue May 16 00:53:30 2017  rustyrazorblade:*with
Tue May 16 00:53:35 2017  rustyrazorblade:probably knows the most
Tue May 16 00:54:05 2017  rustyrazorblade:it would be nice if this could be automated, even if it's a just a nightly thing
Tue May 16 00:54:25 2017  driftx:the problem there is creds
Tue May 16 00:54:31 2017  dikang:Joined the channel
Tue May 16 00:54:37 2017  rustyrazorblade:yeah i have no idea how it's set up
Tue May 16 00:55:08 2017  driftx:I know how I automate my own creds, but that's not really applicable here
Tue May 16 00:56:12 2017  driftx:maybe jfarrell knows of other projects doing something similar?
Tue May 16 00:58:46 2017  jfarrell:setup both mesos and aurora using middleman, but site still in svn which is painful as mentioned
Tue May 16 00:58:57 2017  exlt:I've never stumbled upon the precise versions of sphinx and jekyll for it to cleanly build the exact same html tags, and I've tried many version combinations - I think that's the biggest problem
Tue May 16 01:02:19 2017  exlt:heh s/the biggest/my biggest/ ;) - it really could just be me
Tue May 16 01:04:26 2017  jfarrell:saw this issue with thrift also, tags would get generated differently based on a mac vs ubuntu system
Tue May 16 01:04:49 2017  jfarrell:moved the site build into docker to standardize the setup
Tue May 16 01:18:39 2017  exlt:well, I'll try to revert the things I know it deletes all the time and sift through the rest of this `make add-latest-doc`
Tue May 16 01:18:42 2017  exlt:$ svn status |wc -l
Tue May 16 01:18:45 2017  exlt:684
Tue May 16 01:24:20 2017  driftx:is there no way to like, live in 2017 and not have the site in svn?
Tue May 16 01:24:32 2017  rustyrazorblade:+1 to that
Tue May 16 01:24:39 2017  joaquin:Joined the channel
Tue May 16 01:26:31 2017  rustyrazorblade:if it helps the build process I pinged joaquin, he can help dockerize it. we just need to know the right steps.
Tue May 16 01:27:14 2017  joaquin:i'll be taking off soon, but +1 on rustyrazorblade 's comment. i'd be happy to dockerize that and take away the pain. :)
Tue May 16 01:27:57 2017  driftx:joaquin: oh hey buddy :)
Tue May 16 01:28:15 2017  adamholmberg:Joined the channel
Tue May 16 01:30:40 2017  exlt:this is not working for me, it has removed all of the nodetool sub-pages, so I'm bailing on this commit attempt
Tue May 16 01:31:01 2017  kohlisankalp:Joined the channel
Tue May 16 01:31:21 2017  driftx:pcmanus will fix it I'm sure
Tue May 16 01:31:43 2017  driftx:but, longer term, I'd like something better
Tue May 16 01:31:46 2017  rustyrazorblade:yes
Tue May 16 01:31:48 2017  juiceblender:Joined the channel
Tue May 16 01:32:00 2017  rustyrazorblade:it will be great that he fixes it but someone else needs to understand it
Tue May 16 01:32:11 2017  rustyrazorblade:imagine if it were written down
Tue May 16 01:32:15 2017  rustyrazorblade:IMAGINE
Tue May 16 01:32:45 2017  driftx:I think it was at some point. But those docs are probably outdated too ;)
Tue May 16 01:33:35 2017  rustyrazorblade:heh
Tue May 16 01:33:37 2017  rustyrazorblade:most likely
Tue May 16 01:33:44 2017  exlt:feel my pain.. http://12.am/tmp/doc-update.txt
Tue May 16 01:34:07 2017  rustyrazorblade:i feel it
Tue May 16 01:34:24 2017  exlt::)
Tue May 16 01:34:46 2017  driftx:it's been two lifetimes since I used jekyll or sphinx, and 4 since svn, but I can still feel it.
Tue May 16 01:36:04 2017  exlt:it sort of works, I think, but I didn't want to delete all those publish/doc/latest/tools/nodetool/* pages
Tue May 16 01:36:53 2017  exlt:otherwise, I can live with code hilight tag changes
Tue May 16 01:37:18 2017  exlt:there is a new publish/doc/latest/search.html and js in there, so I do see those generated
Tue May 16 01:39:50 2017  rustyrazorblade:there's a way to generate the nodetool docs
Tue May 16 01:39:56 2017  rustyrazorblade:i reviewed it...
Tue May 16 01:40:48 2017  joaquin:driftx : \o
Tue May 16 01:42:59 2017  exlt:huh, I just ran `ant gen-doc` before heading over to svn, so let me see what a full artifacts build gets me before running the docs build again
Tue May 16 01:43:59 2017  exlt:oh, lookie there.. `[exec] python gen-nodetool-docs.py`
Tue May 16 01:44:52 2017  dikang:Joined the channel
Tue May 16 01:45:10 2017  exlt:ah, shoot, I missed the error at the top of that txt file I posted - Error: Could not find or load main class org.apache.cassandra.tools.NodeTool
Tue May 16 01:45:20 2017  exlt:this'll help, I'm sure
Tue May 16 01:53:24 2017  joaquin:Joined the channel
Tue May 16 01:54:17 2017  rustyrazorblade:you might need to do a full build w/ ant
Tue May 16 01:54:49 2017  exlt:already did - working on sifting the site doc updates now :)
Tue May 16 01:54:59 2017  exlt:that was indeed needed
Tue May 16 01:59:48 2017  agrasso:Joined the channel
Tue May 16 02:00:27 2017  exlt:this is evil, even when it works correctly
Tue May 16 02:00:31 2017  exlt:committed.
Tue May 16 02:03:08 2017  exlt:doing this feels like an icky way to have to do deletes.. svn rm `svn status | egrep '^\!' | awk '{print $2}'`
Tue May 16 02:05:14 2017  exlt:rustyrazorblade: let me know if you see bustage - I see a search and repair doc :)
Tue May 16 02:05:56 2017  rustyrazorblade:it's unfortunate it can't just be something like 1) generate all the docs 2) sync to s3 bucket
Tue May 16 02:06:27 2017  rustyrazorblade:thanks for figuring it out exlt. if you throw it in a jira joaquin said he'd dockerize it
Tue May 16 02:07:49 2017  joaquin:heh, dealt with sphix before and definitely know the adhoc bash commands that have to be executed to make it work. :p yeah, i'll pull together an image with all the dependencies and commands hooked up.
Tue May 16 02:08:17 2017  joaquin:if you can also include any info for how you did it, even if it's not accessible to me, and i'll include that with the scripts and/or companion docs.
Tue May 16 02:08:34 2017  joaquin:i'll try to get it to where anyone with permission just runs one command
Tue May 16 02:09:38 2017  exlt:commit should really be the only thing that requires auth - otherwise, it's just duct tape machetes to your hands and flail at svn ;)
Tue May 16 02:12:17 2017  rustyrazorblade:LOLOLOL
Tue May 16 02:19:52 2017  nickmbailey:Joined the channel
Tue May 16 02:39:07 2017  joaquin:lmao, nice analogy
Tue May 16 02:48:37 2017  agrasso:Joined the channel
Tue May 16 02:55:16 2017  JayZhuang:Joined the channel
Tue May 16 02:59:11 2017  adamholmberg:Joined the channel
Tue May 16 03:13:48 2017  Vijay:Joined the channel
Tue May 16 03:25:04 2017  dikang:Joined the channel
Tue May 16 03:25:49 2017  Vijay:Joined the channel
Tue May 16 03:43:32 2017  agrasso:Joined the channel
Tue May 16 04:09:11 2017  nickmbailey:Joined the channel
Tue May 16 04:18:19 2017  kohlisankalp:Joined the channel
Tue May 16 04:26:03 2017  jfarrell:Joined the channel
Tue May 16 04:29:06 2017  adamholmberg:Joined the channel
Tue May 16 04:34:24 2017  dikang:Joined the channel
Tue May 16 04:53:37 2017  joaquin:Joined the channel
Tue May 16 05:02:36 2017  cosql:Joined the channel
Tue May 16 05:05:18 2017  kohlisankalp:Joined the channel
Tue May 16 05:07:35 2017  kohlisankalp:Joined the channel
Tue May 16 05:37:35 2017  dikang:Joined the channel
Tue May 16 05:49:07 2017  dikang:Joined the channel
Tue May 16 06:42:59 2017  joaquin:Joined the channel
Tue May 16 06:51:26 2017  adamholmberg:Joined the channel
Tue May 16 07:18:25 2017  spodkowinski:Joined the channel
Tue May 16 07:48:35 2017  mebigfatguy:Joined the channel
Tue May 16 07:52:43 2017  adamholmberg:Joined the channel
Tue May 16 07:52:56 2017  cosql:Joined the channel
Tue May 16 07:54:20 2017  cosql:Joined the channel
Tue May 16 07:56:21 2017  kvaster:Joined the channel
Tue May 16 08:22:11 2017  zaller:Joined the channel
Tue May 16 08:27:48 2017  makrusak:Joined the channel
Tue May 16 08:53:06 2017  zaller:Joined the channel
Tue May 16 08:56:37 2017  zaller1:Joined the channel
Tue May 16 08:58:22 2017  zaller:Joined the channel
Tue May 16 09:01:14 2017  agrasso:Joined the channel
Tue May 16 09:54:19 2017  adamholmberg:Joined the channel
Tue May 16 11:10:12 2017  agrasso:Joined the channel
Tue May 16 11:15:36 2017  lqid:Joined the channel
Tue May 16 11:38:19 2017  kvaster:Joined the channel
Tue May 16 11:56:04 2017  adamholmberg:Joined the channel
Tue May 16 11:59:17 2017  adamholmberg:Joined the channel
Tue May 16 12:00:07 2017  adamholmberg:Joined the channel
Tue May 16 12:12:19 2017  agrasso:Joined the channel
Tue May 16 13:13:58 2017  adamholmberg:Joined the channel
Tue May 16 13:25:09 2017  kvaster:Joined the channel
Tue May 16 13:49:42 2017  tjake:jkni: is #13522 a dup?
Tue May 16 13:49:43 2017  CassBotJr:https://issues.apache.org/jira/browse/CASSANDRA-13522 (Open; Unresolved; 2.1.13): "AbstractTracingAwareExecutorService - Uncaught exception on thread - leads to JVM exit"
Tue May 16 13:58:41 2017  clohfink:Joined the channel
Tue May 16 14:09:01 2017  agrasso:Joined the channel
Tue May 16 14:19:34 2017  ynazarov:Joined the channel
Tue May 16 14:33:20 2017  jmckenzie_:Joined the channel
Tue May 16 14:55:14 2017  Vijay:Joined the channel
Tue May 16 15:02:40 2017  kohlisankalp:Joined the channel
Tue May 16 15:15:05 2017  kohlisankalp:Joined the channel
Tue May 16 15:19:26 2017  nickmbailey:Joined the channel
Tue May 16 15:22:22 2017  joaquin:Joined the channel
Tue May 16 15:56:00 2017  jeffj:jmckenzie_: "usability" on #13496 was being kind.
Tue May 16 15:56:01 2017  CassBotJr:https://issues.apache.org/jira/browse/CASSANDRA-13496 (Open; Unresolved; Unscheduled): "table with lots of column ( tens of thousands columns), system is very slow "
Tue May 16 15:56:32 2017  jeffj:tens of thousands of columns, thousands of them being collections, and a lucene 2i on top of it. best data model ever.
Tue May 16 16:13:14 2017  Vijay:Joined the channel
Tue May 16 16:14:07 2017  zaller:Joined the channel
Tue May 16 16:16:07 2017  jmckenzie:Well, it's pretty trivial to pop up a warning in the logs of "You've exceeded N columns. This is probably an anti-pattern - see <link> for more details"
Tue May 16 16:16:18 2017  jmckenzie:Instead of silently degrading / flailing / dying
Tue May 16 16:16:31 2017  jmckenzie:So it's more that usability aspect of things than the "make this work" aspect. :) (jeffj ^)
Tue May 16 16:17:37 2017  jeffj:fair enough
Tue May 16 16:21:16 2017  kvaster:Joined the channel
Tue May 16 16:22:32 2017  cosql:Joined the channel
Tue May 16 16:37:44 2017  adamholmberg:Joined the channel
Tue May 16 16:41:30 2017  adamholm_:Joined the channel
Tue May 16 16:41:43 2017  kohlisankalp:Joined the channel
Tue May 16 16:53:04 2017  clohfink:Joined the channel
Tue May 16 16:55:18 2017  nickmbailey:Joined the channel
Tue May 16 17:03:14 2017  dikang:Joined the channel
Tue May 16 17:06:50 2017  kvaster:Joined the channel
Tue May 16 17:07:47 2017  clohfink:Joined the channel
Tue May 16 17:13:31 2017  kvaster_:Joined the channel
Tue May 16 17:14:49 2017  Vijay:Joined the channel
Tue May 16 17:26:04 2017  adamholmberg:Joined the channel
Tue May 16 17:31:03 2017  gnubeard:Joined the channel
Tue May 16 17:37:18 2017  kvaster_:Joined the channel
Tue May 16 17:47:01 2017  yukim:Joined the channel
Tue May 16 17:53:23 2017  nickmbailey:Joined the channel
Tue May 16 17:55:27 2017  kohlisan_:Joined the channel
Tue May 16 18:17:00 2017  kvaster_:Joined the channel
Tue May 16 18:20:15 2017  nickmbailey:Joined the channel
Tue May 16 18:20:34 2017  ynazarov:Joined the channel
Tue May 16 18:38:10 2017  kvaster_:Joined the channel
Tue May 16 18:58:39 2017  bbromhead:Joined the channel
Tue May 16 19:07:50 2017  agrasso:Joined the channel
Tue May 16 19:10:54 2017  jmckenzie_:Joined the channel
Tue May 16 19:25:56 2017  clohfink:Joined the channel
Tue May 16 19:32:01 2017  rustyrazorblade:anyone want to review this patch that makes stress a little bit less terrible WRT batches? https://issues.apache.org/jira/browse/CASSANDRA-11105
Tue May 16 19:32:10 2017  agrasso:Joined the channel
Tue May 16 19:55:26 2017  zaller:Joined the channel
Tue May 16 19:56:42 2017  dikang:Joined the channel
Tue May 16 20:03:26 2017  bbromhead:Joined the channel
Tue May 16 20:10:28 2017  ostefano_:Hi, I have a question re the reading path in C*, in particular about range tombstones.
Tue May 16 20:10:38 2017  ostefano_:I was under the impression that the information stored within a range tombstone was used to seek within the sstables (so to avoid scans).
Tue May 16 20:10:44 2017  ostefano_:However, based on what I read from the source code, the liveness of a cell is reconciled while iterating over the row.
Tue May 16 20:10:47 2017  ostefano_:Is that correct?
Tue May 16 20:11:27 2017  ostefano_:It's my first time diving into C* source code (3.0) so I admit I might be missing something here
Tue May 16 20:20:09 2017  kohlisankalp:Joined the channel
Tue May 16 21:00:57 2017  spodkowinski:Joined the channel
Tue May 16 21:01:03 2017  bbromhead:Joined the channel
Tue May 16 21:52:52 2017  EdwardCapriolo:Joined the channel
Tue May 16 21:53:53 2017  kohlisankalp:Joined the channel
Tue May 16 22:16:48 2017  kohlisankalp:Joined the channel
Tue May 16 22:30:28 2017  kohlisankalp:Joined the channel
Tue May 16 22:42:15 2017  zaller:Joined the channel
Tue May 16 22:52:51 2017  zaller:Joined the channel
Tue May 16 23:07:53 2017  kohlisankalp:Joined the channel
Tue May 16 23:16:10 2017  EdwardCapriolo:Joined the channel
Tue May 16 23:29:56 2017  kohlisankalp:Joined the channel
Tue May 16 23:44:53 2017  bbromhead:Joined the channel
Tue May 16 23:51:05 2017  clohfink:Joined the channel
Tue May 16 23:53:51 2017  clohfink:Joined the channel

Comments