Re: Testing Upgrade to 1.6.x
On Sun, Apr 27, 2014 at 1:38 AM, Robert Samuel Newson rnew...@apache.orgwrote: Did you 'make clean' first? This looks like you have a module with the old #changes_args record and other modules with the new one. I ran 'make distclean' before ./boostrap and ./configure
Testing Upgrade to 1.6.x
Hi, As I see that 1.6.x is coming along, I've tested upgrading an 1.5 db to the current 1.6.x branch (26cff2d51473cbc8c0dfc655b2dbb986493bbcb8). I've stopped couch, compiled 1.6.x with R15B03 and installed the result on an Ubuntu 12.04.4 LTS. The instance always failed with the logging output shown in https://gist.github.com/stefankoegl/11332148 Is this a bug in the 1.6.x branch, or is my upgrade process somehow broken? -- Stefan
Re: [ANN] Dave Cottlehuber joins the PMC
Congratulations! Well deserved, Dave \°/ On Tue, Oct 30, 2012 at 11:55 PM, Noah Slater nsla...@apache.org wrote: Hey folks, I am delighted to announce that Dave Cottlehuber joins the Apache CouchDB Project Management Committee today. Dave has made outstanding, sustained contributions, and this appointment is an official acknowledgement of his position within the community, and our trust in his ability to provide oversight for the project. Everybody, please join me in congratulating Dave! I would also like to take this opportunity to bid a fond farewell to J. Chris Anderson, who has resigned from the PMC. J. Chris was instrumental to much of CouchDB's early success, and we wish him all the best with his future projects. Thanks, -- NS
Re: COUCHDB-1444 (missing_named_view error on existing javascript design doc and view)
Hi, On Fri, Aug 3, 2012 at 1:50 PM, Robert Newson robert.new...@gmail.com wrote: Yes, I've been looking into it but I can't induce the condition locally, which hinders my progress considerably. I have made a related fix on couchdb master (ce7204b7eb64ac98d4445130fc4e647ed5181da9) which might have fixed this but I can't confirm. If you are able to induce this reliably, I would like to work with you to resolve this issue. IRC is best. Unfortunately I can't reproduce this reliably. The problem only appears on one of my production systems, but there rather often (at least every 2nd day, sometimes even 5 times a day). However, if you could prepare a patch against 1.2.0, I'd try to patch this system and see what happens. -- Stefan On 2 Aug 2012, at 22:15, Stefan Kögl wrote: Hi, I just wanted to ask if any of the devs would be willing to help with tracking down and solving COUCHDB-1444 (missing_named_view error on existing javascript design doc and view), which has been open since March. Already 7 people who were affected by this issue (some on production systems) have commented on the issue, confirming it for combinations of 1.1.1, 1.2.0, 1.3.0@master (f0d6f19bc8) and R13B03, R14, R15B01, but apparently no one could provide any useful debugging output. I lack the Erlang skills to tackle this issue myself, but I'd like to help with debugging as much as I can. Any feedback, thoughts, etc would be very much appreciated! Thanks, -- Stefan
COUCHDB-1444 (missing_named_view error on existing javascript design doc and view)
Hi, I just wanted to ask if any of the devs would be willing to help with tracking down and solving COUCHDB-1444 (missing_named_view error on existing javascript design doc and view), which has been open since March. Already 7 people who were affected by this issue (some on production systems) have commented on the issue, confirming it for combinations of 1.1.1, 1.2.0, 1.3.0@master (f0d6f19bc8) and R13B03, R14, R15B01, but apparently no one could provide any useful debugging output. I lack the Erlang skills to tackle this issue myself, but I'd like to help with debugging as much as I can. Any feedback, thoughts, etc would be very much appreciated! Thanks, -- Stefan
Re: [VOTE] Apache CouchDB 1.2.0 release, third round
Hi everybody, I just wanted to raise some attention to the DB compaction bug discovered today https://issues.apache.org/jira/browse/COUCHDB-1451 While I initially discovered the bug with a 1.1.2 instance, this could also affect 1.2. I think this issue should be resolved before closing the vote. Filipe already provided a patch, so it shouldn't take too long. -- Stefan
Re: [VOTE] Apache CouchDB 1.2.0 release, third round
On 03/28/2012 07:59 PM, Jan Lehnardt wrote: Hi Stefan, thanks for bringing this up. This is just to explain the procedure. A vote happens on a fixed package and no changes to the package can be made during the vote. If we find a flaw significant enough to include a fix into the release that is currently being voted on, we can abort the vote, fix, re-package and start a new vote, like we've done before. Thanks for the explanation, that's exactly what I expected ;) What I meant was, that we should either * rule out that this affects 1.2.x (but apparently it does) * fix it and start another round of votes Anyway, I wanted to be in time before somebody closes the vote and 1.2 might be shipped with a bug. -- Stefan
Re: [VOTE] Apache CouchDB 1.2.0 release, third round
On 03/28/2012 08:08 PM, Robert Newson wrote: Can you clarify the conditions under which this bug occurs? I'm inclined to agree with Filipe that it's release blocking, but if it the conditions to induce are very rare, I might change my mind. Given that you induced it without apparent effort, I don't think it's likely to be rare. I have a setup where I redirect live read traffic to test instances, while replicating writes from a stable master. This is the first such incident in a few weeks of doing this (with maybe 5 to 10 db compactions), so I can't judge how often it occurs. From my point of view there was no special work load on the instance when it happend. -- Stefan
Re: [VOTE] Apache CouchDB 1.2.0 release, third round
On Sun, Mar 25, 2012 at 1:04 PM, Paul Davis paul.joseph.da...@gmail.com wrote: Try adding: --with-js-lib-name=mozjs185 to your ./configure invocation That worked. +1 from me :) Also, why is libjs installed on this system? Was that a custom thing you did or is it included for some reason in a package? I can't remember why, tbh. Is that important somehow? -- Stefan
Re: [VOTE] Apache CouchDB 1.2.0 release, third round
Hi, On Sat, Mar 24, 2012 at 6:39 PM, Noah Slater nsla...@tumbolia.org wrote: Please follow the test procedure before voting: http://wiki.apache.org/couchdb/Test_procedure Its probably related to my system (Ubuntu precise beta 1 with current updates), but I'm getting the following errors during make check (in the make part). http://friendpaste.com/3cmPcgwU1hXHYjXhixaLFk Any ideas whether I need to fix my system or its a bug? -- Stefan
Re: [VOTE] Apache CouchDB 1.2.0 release, third round
Hi, On Sun, Mar 25, 2012 at 9:27 AM, Paul Davis paul.joseph.da...@gmail.com wrote: Also the contents of mozjs185.pc This is probably in either /usr/lib/pkconfig or /usr/local/lib/pkgconfig. http://friendpaste.com/1UJ8uaGpYJRSpfhVZqrNvF Can you paste me the contents of config.log? http://friendpaste.com/1AXWuMuTUzCMCXePuSJijA -- Stefan
Re: {error,emfile} on CouchDB 1.2.x
On Mon, Mar 19, 2012 at 9:31 AM, Randall Leeds randall.le...@gmail.com wrote: Fixed on 1.2.x and 1.1.x. Need to sleep and take a look at how I want to handle it on master. Thanks again for picking up on this one, Stefan. It's been in there since forever and I'd definitely seen the symptom without knowing the cause. Thanks for taking care of this so quickly -- looking forward to a great release :) -- Stefan
{error,emfile} on CouchDB 1.2.x
Hi, Another thing I noticed during my tests of CouchDB 1.2.x. I redirected live traffic to the instance and after a rather short time, requests were failing with the following information in the logs: [Sun, 18 Mar 2012 16:39:24 GMT] [error] [0.27554.2] {error_report,0.31.0, {0.27554.2,std_error, [{application,mochiweb}, Accept failed error, {error,emfile}]}} [Sun, 18 Mar 2012 16:39:24 GMT] [error] [0.27554.2] {error_report,0.31.0, {0.27554.2,crash_report, [[{initial_call, {mochiweb_acceptor,init, ['Argument__1','Argument__2', 'Argument__3']}}, {pid,0.27554.2}, {registered_name,[]}, {error_info, {exit, {error,accept_failed}, [{mochiweb_acceptor,init,3}, {proc_lib,init_p_do_apply,3}]}}, {ancestors, [couch_httpd,couch_secondary_services, couch_server_sup,0.32.0]}, {messages,[]}, {links,[0.129.0]}, {dictionary,[]}, {trap_exit,false}, {status,running}, {heap_size,233}, {stack_size,24}, {reductions,244}], []]}} I think emfile means that CouchDB (or mochiweb?) couldn't open any more files / connections. I've set the (hard and soft) nofile limit for user couchdb to 4096, but didn't raise the ERL_MAX_PORTS accordingly. Anyway, as soon as the error occured, CouchDB started writing most of my view files from scratch, rendering the instance unusable. I'd expect CouchDB to fail more gracefully when the maximum number of open files is reached. Is this a bug or expected behaviour? -- Stefan
Re: Update Conflict for PUT/DELETE in _replicator
In the meantime I tried copying the _replicator database to another instance, where I could delete the entry without problems. However it still doesn't work on the initial instance. If one of the committers is interested, I could organize either remote access via HTTP, or shell access to the machine it is running on. -- Stefan On Fri, Mar 2, 2012 at 4:30 PM, Stefan Kögl koeglste...@gmail.com wrote: On Fri, Mar 2, 2012 at 4:06 PM, Jan Lehnardt j...@apache.org wrote: I just created a replication doc under 1.1.1 and then copied the _replicator.couch file to a 1.2.x. On update I the expected result Robert also got (Only the replicator can edit replication documents that are in the triggered state.. a curl -X DELETE on the doc with ?rev=4-abcd... (no quotes) also worked. The document was created with 1.2.x, from around the time of the second RC. I also tried with quotes and got $ curl -sv -X DELETE http://stefan:*@127.0.0.1:5984/_replicator/mygpo?rev=\131-57b4da8d3163468cb0bbf4fd30c87832\; * About to connect() to 127.0.0.1 port 5984 (#0) * Trying 127.0.0.1... connected * Connected to 127.0.0.1 (127.0.0.1) port 5984 (#0) * Server auth using Basic with user 'stefan' DELETE /_replicator/mygpo?rev=131-57b4da8d3163468cb0bbf4fd30c87832 HTTP/1.1 Authorization: Basic ** User-Agent: curl/7.19.7 (x86_64-pc-linux-gnu) libcurl/7.19.7 OpenSSL/0.9.8k zlib/1.2.3.3 libidn/1.15 Host: 127.0.0.1:5984 Accept: */* HTTP/1.1 500 Internal Server Error Server: CouchDB/1.2.0 (Erlang OTP/R14B04) Date: Fri, 02 Mar 2012 15:18:31 GMT Content-Type: text/plain; charset=utf-8 Content-Length: 44 Cache-Control: must-revalidate {error:unknown_error,reason:badarg} * Connection #0 to host 127.0.0.1 left intact * Closing connection #0 After that I also tried compacting the _replicator database, but also that didn't change anything. -- Stefan
Re: Update Conflict for PUT/DELETE in _replicator
On Mon, Mar 12, 2012 at 2:01 PM, Jason Smith j...@iriscouch.com wrote: On Mon, Mar 12, 2012 at 12:57 PM, Stefan Kögl koeglste...@gmail.com wrote: In the meantime I tried copying the _replicator database to another instance, where I could delete the entry without problems. However it still doesn't work on the initial instance. If one of the committers is interested, I could organize either remote access via HTTP, or shell access to the machine it is running on. Hm, if you copied the _replicator.couch file from the 1.2.x prerelease to version 1.1 then it will not support the newer file format. You could replicate to it, or since replication docs have no attachments, just query _all_docs, massage that into a _bulk_docs, and post it to the other couch. That's not the problem. I copied the database to another 1.2.x CouchDB, which could read and update it correctly. The problem is the original 1.2.x instance (both are on the current 1.2.x branch, btw) where I created the entry but can not update / delete it anymore. -- Stefan
Re: Update Conflict for PUT/DELETE in _replicator
On Mon, Mar 12, 2012 at 2:20 PM, Robert Newson rnew...@apache.org wrote: I'd welcome a chance to access this database. couchdb admin level access is sufficient for now, email me directly at rnew...@apache.org. For general information: The problem spontaneously disappeared when I tried to reproduce it with the admin I created for rnewson... I'll report back in case it happens again. -- Stefan
Re: Crash of CouchDB 1.2.x
Hi, I had my CouchDB 1.2.x (fb72251bc7114b07f0667867226ec9e200732dac) crash again twice today. The first one was while the instance was pull replicating (which failed due to the source being unreachable), and compacting a rather large view (from ~216G disk size to about 57G data size, if that's relevant). Here's the log that shows the crash http://friendpaste.com/41Idie3gGdQRxJPEyVHpTR After the crash the view compaction stopped, and I tried to restart it $ curl -H Content-Type: application/json -X POST http://stefan:@localhost:5984/mygpo/_compact/users-tmp {error:timeout,reason:{gen_server,call,\n [0.19783.69,\n {start_compact,#Funcouch_view_compactor.0.15011741}]}} http://friendpaste.com/2A086gHN8dNEJHPpMkDrPO I assume this is because deleting the .compact.view file took too long. The compaction started anyway, though. Besides the replication, there were no other activities on the server. Please let me know if I can assist with debugging somehow. -- Stefan On Fri, Mar 2, 2012 at 11:51 AM, Jan Lehnardt j...@apache.org wrote: On Mar 2, 2012, at 11:29 , Stefan Kögl wrote: On Thu, Mar 1, 2012 at 9:39 PM, Jan Lehnardt j...@apache.org wrote: Where in there did you do the git pull? And was a make clean or git clean involved? IIRC I did not pull in between, only apply the patch I mentioned earlier ( http://friendpaste.com/178nPFgfyyeGf2vtNRpL0w ). And I probably did a make clean make make install. I think you should be in the clear with that procedure, but just to be sure, I think it'd be worth rm'ing all .beam files you find manually after the uninstall. Done, I'll report back if the problem appears again. Thanks Stefan, your help here is really appreciated :) Cheers Jan --
Re: Crash of CouchDB 1.2.x
On 03/11/2012 01:33 PM, Bob Dionne wrote: At a glance I would suspect you've run out of disk space and the error thrown is not caught resulting in the badmatch, At the time of the crash there were about 70G of free space left, which is enough for the compaction to finish. -- Stefan
Re: Crash of CouchDB 1.2.x
On 03/11/2012 02:32 PM, Jason Smith wrote: Longshot, but is it possible that couch had a file handle to an unlinked file, so once the (OS) process crashed, the space was freed? Hmm.. that might be possible. I ran a database compaction before that. When I noticed the crash I saw that the db compaction finished, but it might be possible that it still had a handle to the old db file. How should we proceed from here? Is it possible for me to provide further information about that in retrospect? -- Stefan
Re: Crash of CouchDB 1.2.x
On Thu, Mar 1, 2012 at 9:39 PM, Jan Lehnardt j...@apache.org wrote: Where in there did you do the git pull? And was a make clean or git clean involved? IIRC I did not pull in between, only apply the patch I mentioned earlier ( http://friendpaste.com/178nPFgfyyeGf2vtNRpL0w ). And I probably did a make clean make make install. I think you should be in the clear with that procedure, but just to be sure, I think it'd be worth rm'ing all .beam files you find manually after the uninstall. Done, I'll report back if the problem appears again. -- Stefan
Re: Update Conflict for PUT/DELETE in _replicator
On Fri, Mar 2, 2012 at 1:39 PM, Jan Lehnardt j...@apache.org wrote: On Mar 2, 2012, at 13:25 , Stefan Kögl wrote: Again something I noticed during my 1.2.x experiments: It seems I can't update or remove a document from the _replicator database, which I use for pull-replication into my 1.2.x instance. # get current _rev $ curl http://127.0.0.1:5984/_replicator/mygpo {_id:mygpo,_rev:131-57b4da8d3163468cb0bbf4fd30c87832,source:,target:http://127.0.0.1:5984/mygpo,create_target:false,continuous:true,user_ctx:{name:stefan,roles:[admin]},owner:stefan,_replication_state:triggered,_replication_state_time:2012-03-02T02:56:12+00:00,_replication_id:f9fc5457b278d3cdb1ba2f1881253b04} # try to delete $ curl -X DELETE http://127.0.0.1:5984/_replicator/mygpo?_rev=131-57b4da8d3163468cb0bbf4fd30c87832; {error:conflict,reason:Document update conflict.} this should be ?rev=... (no underscore) Of course, I also tried that but it didn't work either... see below. $ curl -X PUT http://127.0.0.1:5984/_replicator/mygpo; -d @replication.json Can you try adding the ?rev= in the URL as well? # get the rev first $ curl http://127.0.0.1:5984/_replicator/mygpo {_id:mygpo,_rev:131-57b4da8d3163468cb0bbf4fd30c87832,source:,target:http://127.0.0.1:5984/mygpo,create_target:false,continuous:true,user_ctx:{name:stefan,roles:[admin]},owner:stefan,_replication_state:triggered,_replication_state_time:2012-03-02T02:56:12+00:00,_replication_id:f9fc5457b278d3cdb1ba2f1881253b04} # delete with rev instead of _rev $ curl -X DELETE http://127.0.0.1:5984/_replicator/mygpo?rev=131-57b4da8d3163468cb0bbf4fd30c87832; {error:conflict,reason:Document update conflict.} # now PUT with rev in URL $ curl -X PUT http://127.0.0.1:5984/_replicator/mygpo?rev=131-57b4da8d3163468cb0bbf4fd30c87832; -d @replication.json {error:conflict,reason:Document update conflict.} # still the same rev (no other change in between) $ curl http://127.0.0.1:5984/_replicator/mygpo {_id:mygpo,_rev:131-57b4da8d3163468cb0bbf4fd30c87832,source:,target:http://127.0.0.1:5984/mygpo,create_target:false,continuous:true,user_ctx:{name:stefan,roles:[admin]},owner:stefan,_replication_state:triggered,_replication_state_time:2012-03-02T02:56:12+00:00,_replication_id:f9fc5457b278d3cdb1ba2f1881253b04} -- Stefan
Re: Update Conflict for PUT/DELETE in _replicator
On Fri, Mar 2, 2012 at 2:54 PM, Robert Newson rnew...@apache.org wrote: could you redo the DELETE or PUT with '-sv' so we can see what's really being sent? Perhaps there's a weird shell thing happening causing the rev to be sent incorrectly. $ curl -sv -X DELETE http://stefan:*@127.0.0.1:5984/_replicator/mygpo?rev=131-57b4da8d3163468cb0bbf4fd30c87832; * About to connect() to 127.0.0.1 port 5984 (#0) * Trying 127.0.0.1... connected * Connected to 127.0.0.1 (127.0.0.1) port 5984 (#0) * Server auth using Basic with user 'stefan' DELETE /_replicator/mygpo?rev=131-57b4da8d3163468cb0bbf4fd30c87832 HTTP/1.1 Authorization: Basic ** User-Agent: curl/7.19.7 (x86_64-pc-linux-gnu) libcurl/7.19.7 OpenSSL/0.9.8k zlib/1.2.3.3 libidn/1.15 Host: 127.0.0.1:5984 Accept: */* HTTP/1.1 409 Conflict Server: CouchDB/1.2.0 (Erlang OTP/R14B04) Date: Fri, 02 Mar 2012 13:57:01 GMT Content-Type: text/plain; charset=utf-8 Content-Length: 58 Cache-Control: must-revalidate {error:conflict,reason:Document update conflict.} * Connection #0 to host 127.0.0.1 left intact * Closing connection #0 I verified the rev before and after the DELETE to make sure it hasn't changed in between. On 2 March 2012 13:51, Robert Newson rnew...@apache.org wrote: tbh you should be getting this error: 'Only the replicator can edit replication documents that are in the triggered state'. I do. Shouldn't I be able to DELETE the replication document? -- Stefan
Re: Update Conflict for PUT/DELETE in _replicator
On Fri, Mar 2, 2012 at 4:06 PM, Jan Lehnardt j...@apache.org wrote: I just created a replication doc under 1.1.1 and then copied the _replicator.couch file to a 1.2.x. On update I the expected result Robert also got (Only the replicator can edit replication documents that are in the triggered state.. a curl -X DELETE on the doc with ?rev=4-abcd... (no quotes) also worked. The document was created with 1.2.x, from around the time of the second RC. I also tried with quotes and got $ curl -sv -X DELETE http://stefan:*@127.0.0.1:5984/_replicator/mygpo?rev=\131-57b4da8d3163468cb0bbf4fd30c87832\; * About to connect() to 127.0.0.1 port 5984 (#0) * Trying 127.0.0.1... connected * Connected to 127.0.0.1 (127.0.0.1) port 5984 (#0) * Server auth using Basic with user 'stefan' DELETE /_replicator/mygpo?rev=131-57b4da8d3163468cb0bbf4fd30c87832 HTTP/1.1 Authorization: Basic ** User-Agent: curl/7.19.7 (x86_64-pc-linux-gnu) libcurl/7.19.7 OpenSSL/0.9.8k zlib/1.2.3.3 libidn/1.15 Host: 127.0.0.1:5984 Accept: */* HTTP/1.1 500 Internal Server Error Server: CouchDB/1.2.0 (Erlang OTP/R14B04) Date: Fri, 02 Mar 2012 15:18:31 GMT Content-Type: text/plain; charset=utf-8 Content-Length: 44 Cache-Control: must-revalidate {error:unknown_error,reason:badarg} * Connection #0 to host 127.0.0.1 left intact * Closing connection #0 After that I also tried compacting the _replicator database, but also that didn't change anything. -- Stefan
Crash of CouchDB 1.2.x
Hello, My experiments to replicate some live data / traffic to a CouchDB 1.2.x (running the current 1.2.x branch + the patch from [1]) that sparked the indexing speed discussions, did also yield another (potential) problem. First sorry for not further reporting back any performance measurements, but I didn't yet find the time to run the tests on my machines. Anyway, I found the following stack traces in my log (after noticing that some requests failed and compaction of a view stopped) http://skoegl.net/~stefan/tmp/couchdb-1.2.x-crash.txt The files starts at the first failed requests. Every request before that returned a positiv (ie 2xx) status code. The crash might have some natural reason (such as timeouts, lack of RAM, etc), but I'm not sure how to interpret Erlang stack traces. Can somebody point me in the right direction for diagnosing the problem? Thanks, -- Stefan [1] http://friendpaste.com/178nPFgfyyeGf2vtNRpL0w
Re: Crash of CouchDB 1.2.x
On Thu, Mar 1, 2012 at 4:52 PM, Jan Lehnardt j...@apache.org wrote: Can you tell us how you installed 1.2.x? Is it a fresh installation, did you do an in-place update from an earlier installation (earlier 1.2.x or 1.1.x or 1.0.x? I first did a fresh install of 1.2.x using R15B. I then removed R15B, installed R14B04 (both from source), compiled 1.2.x with the patch I mentioned earlier, and did an in-place update. If this is a problem, I could remove CouchDB first and do a fresh install instead. What would be the preferred way to do a clean uninstall? -- Stefan On Mar 1, 2012, at 12:17 , Stefan Kögl wrote: Hello, My experiments to replicate some live data / traffic to a CouchDB 1.2.x (running the current 1.2.x branch + the patch from [1]) that sparked the indexing speed discussions, did also yield another (potential) problem. First sorry for not further reporting back any performance measurements, but I didn't yet find the time to run the tests on my machines. Anyway, I found the following stack traces in my log (after noticing that some requests failed and compaction of a view stopped) http://skoegl.net/~stefan/tmp/couchdb-1.2.x-crash.txt The files starts at the first failed requests. Every request before that returned a positiv (ie 2xx) status code. The crash might have some natural reason (such as timeouts, lack of RAM, etc), but I'm not sure how to interpret Erlang stack traces. Can somebody point me in the right direction for diagnosing the problem? Thanks, -- Stefan [1] http://friendpaste.com/178nPFgfyyeGf2vtNRpL0w
Re: Crash of CouchDB 1.2.x
On 03/01/2012 07:38 PM, Jan Lehnardt wrote: On Mar 1, 2012, at 19:18 , Stefan Kögl wrote: If this is a problem, I could remove CouchDB first and do a fresh install instead. What would be the preferred way to do a clean uninstall? I don't want to claim that this is definitely the cause for your problem, but it'd be great if you could do a clean, fresh, empty install to make sure we can rule that out as :) I just did /etc/init.d/couchdb stop make uninstall make install # edit local.ini -- why does that get removed anyway? /etc/init.d/couchdb start Is that enough to count as a fresh install, or should I do anything else? I'll continue monitoring the instance. Previously the error happened after a few days, so I can't say yet if the re-install changed anything. -- Stefan
Re: Crash of CouchDB 1.2.x
Hi, On Fri, Mar 2, 2012 at 4:33 AM, Nathan Vander Wilt nate-li...@calftrail.com wrote: Was your server under heavy load? Did you end up with a bunch of zombie couchjs processes? The crash occured under load, but there are no zombies - at least not anymore. If the crash happens again I'll try to inspect it more closely. -- Stefan
Re: couchdb conflict resolution
On Tue, Feb 15, 2011 at 5:14 PM, Aaron Boxer boxe...@gmail.com wrote: I am very interested in understanding how conflict resolution works in couchdb: Is there a technical overview, somewhere, of how a node decides which revision wins? After a conflict is resolved, are old revisions discarded? Any technical details, short of slogging through the code, would be very welcome. I'm not sure if this is the right level of detail for your purposes, but I found http://guide.couchdb.org/draft/conflicts.html very interesting. -- Stefan