PMXBOT Log file Viewer

Help | Karma | Search:

#mongodb logs for Friday the 12th of February, 2016

(Back to #mongodb overview) (Back to channel listing) (Animate logs)
[00:03:17] <GothAlice> Derick: ^ ?
[00:11:00] <Derick> chalker_: sometimes - probably not at 1am
[00:11:07] <Derick> chalker_: better email...
[00:11:35] <chalker_> Derick: Thanks. Is he using the same nickname in IRC?
[00:12:22] <Derick> to be fair, I haven't seen him on here
[00:12:29] <Derick> better email
[00:12:49] <chalker_> Ah, ok. Thanks again!
[06:52:39] <sabrehagen> how do i restructure this logic using a query to reduce my number of database accesses? i have a query that returns documents in a certain time range. each document has a device id. for each of those documents i want to see if there is a document matching
[06:53:10] <sabrehagen> that device id in the 10 seconds prior. the initial number of documents can be 40,000+. my logic then sends 40,000 database queries to check the second piece of criteria.
[06:53:28] <sabrehagen> without the aggregation framework, how would i minimise the number of queries?
[06:59:36] <Waheedi> i have added a 4th mongodb node to a replicateset. I'm not sure why i see two primaries in the replicaset
[06:59:58] <Waheedi> while the added server has been added using normal rs.add operation
[07:00:09] <joannac> Waheedi: o.O output of rs.status() please
[07:01:54] <Waheedi> https://gist.github.com/dakwak/413bad17bafb7892a445
[07:02:24] <Waheedi> joannac: the .nl.dakwak.com servers are located in the Netherlands and the other two servers are in the US
[07:02:53] <joannac> what about rs.status() on the 2 machines claiming to be primary?
[07:03:12] <Waheedi> the real primary is mongodb.nl.dakwak.com
[07:03:20] <Waheedi> the fake one is mongodb2.nl.dakwak.com
[07:03:35] <Waheedi> when u rs.status from mongodb2.nl.dakwak.com u see it says its a secondary
[07:03:44] <Waheedi> for self
[07:03:54] <joannac> what version of mongodb?
[07:04:04] <Waheedi> 2.4.14
[07:04:44] <joannac> weird. but there's not actually 2 primaries, right?
[07:04:52] <joannac> do any of the other servers think there's 2 primaries?
[07:04:57] <Waheedi> actually there is for some clients
[07:05:14] <Waheedi> some clients can't write on the db anymore
[07:05:32] <Waheedi> two servers think there is two primaries
[07:05:38] <joannac> which 2?
[07:05:44] <Waheedi> not one, both dakwak-mongo-6 and mongo4
[07:06:29] <joannac> what happens if you restart one of those 2?
[07:06:54] <Waheedi> https://gist.github.com/dakwak/88faadac6d6d9ae82978
[07:07:19] <Waheedi> the primaries or secondaries ?
[07:07:38] <joannac> one of the 2 hosts that has the wrong idea of the replica set
[07:07:44] <joannac> so one of the secondaries, I think
[07:07:52] <Waheedi> ok let me try
[07:08:23] <Waheedi> still thinks its has two primaries
[07:08:31] <Waheedi> mongodb start/running, process 699
[07:08:57] <Waheedi> ok
[07:08:59] <Waheedi> not now
[07:09:13] <joannac> weird...
[07:09:20] <Waheedi> this is crazy now, wait
[07:09:44] <Waheedi> https://gist.github.com/dakwak/159faa7265d689dcaaf4
[07:10:18] <joannac> on dakwak-mongo4
[07:10:34] <joannac> can you ping mongodb.nl.dakwak.com and mongodb2.nl.dakwak.com ?
[07:10:51] <Waheedi> two primaries on mongo-4
[07:11:28] <joannac> do they by chance map to the same IP?
[07:13:10] <Waheedi> u mean mongodb.nl.dakwak.com and mongodb2.nl.dakwak.com?
[07:13:38] <joannac> isn't that what I said?
[07:13:50] <Waheedi> from dakwak-mongo6:- ping mongodb2.nl.dakwak.com
[07:13:51] <Waheedi> PING mongodb2.nl.dakwak.com (5.153.56.2)
[07:14:08] <Waheedi> akwak-mongo6:~# ping mongodb.nl.dakwak.com
[07:14:08] <Waheedi> PING mongodb.nl.dakwak.com (159.8.30.116) 56(84) bytes of data.
[07:14:21] <joannac> is that a host that is seeing 2 primaries?
[07:14:25] <Waheedi> the same results from dakwak-mongo4
[07:14:26] <Waheedi> yes
[07:14:45] <joannac> okay, that is super weird
[07:14:50] <Waheedi> dakwak-mongo6 is a bit moody and decided to show no primaries for a while
[07:15:22] <Waheedi> now it shows couple of them
[07:15:40] <Waheedi> my guess is that mongodb2.nl has a corrupt local database
[07:15:54] <joannac> what does db.isMaster() show on dakwak-mongo4 or dakwak-mongo6
[07:16:16] <Waheedi> "primary" : "mongodb.nl.dakwak.com:27017",
[07:16:23] <Waheedi> not two hosts at least
[07:17:10] <Waheedi> is there a way to rsync local database for one of host?
[07:17:17] <Waheedi> resync
[07:17:49] <joannac> no
[07:18:05] <joannac> and I don't know why you think the problem is with mongodb2.nl
[07:18:35] <Waheedi> Its the most recent node and there is no communication issues between all nodes for sure 100%
[07:19:57] <Waheedi> do u think i need to remove that node from the rs and put it back ?
[07:20:22] <joannac> no
[07:20:46] <joannac> Can you do a complete hostname check on all hosts?
[07:20:56] <Waheedi> give me a command
[07:20:57] <joannac> on each host, "getent hosts HOSTNAME"
[07:21:06] <joannac> replace HOSTNAME with each of the 4 hostnames
[07:21:07] <Waheedi> alrights
[07:22:17] <Waheedi> i found something interesting on the first node
[07:22:20] <Waheedi> was dakwak-mongo6
[07:22:42] <lowbro> https://goo.gl/lP11uV
[07:22:47] <lowbro> are those webinars free of charge?
[07:23:18] <Waheedi> joannac: I think its dakwak-mongo6 not being able to reach any of the .nl servers
[07:23:47] <Waheedi> that is super weird for me now
[07:31:04] <Waheedi> ok i will paste the result anyway :)
[07:34:59] <Waheedi> Thank you joannac
[07:59:27] <Waheedi> joannac: https://gist.github.com/dakwak/cb5624445f68e7972d99
[07:59:35] <Waheedi> still showing two primaries
[08:02:33] <Waheedi> the other thing was, Im not sure why nearest read preference on the query level does not apply the primary as an option to read from
[08:02:48] <Waheedi> I though nearest would be regardless of role
[08:02:56] <Waheedi> thought*
[08:19:58] <Keksike> how can I figure out why this kind of query is taking such a long time https://gist.github.com/Keksike/517341c5cd1c861ed0bd
[08:20:10] <Keksike> that is output from db.system.profile.find() after setting a 100ms profiling time
[08:26:36] <Ceunincksken> Kaksike: Try running your query with 'explain()'.
[08:27:00] <Ceunincksken> Kaksike: Thus: db.<collection>.find(<query>).explain();
[08:29:46] <Ceunincksken> Kaksike: Can you post your query and some sample data?
[08:36:41] <Keksike> Ceunincksken: I tried running the same exact query with runCommand (you cant use explain with distinct) and it gave me https://gist.github.com/Keksike/517341c5cd1c861ed0bd the upper file
[08:36:48] <Keksike> so it runs perfectly fine if I do it manually
[08:36:59] <Keksike> but when the program does it, its a different situation and runs much slower
[08:39:45] <Ceunincksken> Keksike: But you're saying that on the database it runs fine but from within your application it's slow, is that correct?
[08:40:51] <Keksike> well, yes, if I just do it directly in the mongo shell, it runs in 0ms
[08:41:05] <Ceunincksken> Weird.
[08:41:16] <Keksike> but my program logs that 181ms into system.profile
[08:41:26] <Keksike> can it be related to indexes or locks?
[08:41:30] <Ceunincksken> Are you sure that you're executing the exact same query?
[08:41:53] <Keksike> well you can check from the gist
[08:41:59] <Keksike> it should be doing the same exact thing
[08:42:42] <Ceunincksken> Sorry but I'm affraid I can't help than.
[08:43:27] <Keksike> ok :)
[08:43:30] <Keksike> thanks anyways
[08:43:35] <Keksike> ill try and figure it out
[09:43:31] <pchoo> Morning all, I have a question about matching on an array of sub documents - if we take the example here: https://docs.mongodb.org/v3.0/tutorial/query-documents/#specify-multiple-criteria-for-array-elements - how can I find all documents which have a memo from shipping, but not from billing?
[09:47:13] <pchoo> This is for use in an aggregation pipeline, would I be better off unwinding the array of documents, then grouping the information I need, and using $match to find those where billing doesn't exist?
[09:47:59] <Ceunincksken> That would probably be the best solution.
[09:51:19] <pchoo> thanks, I'll investigate that route
[09:59:53] <tantamount> Is there any way to truncate the length of an array?
[09:59:59] <tantamount> During an aggregation projection
[10:11:20] <Zelest> https://www.reddit.com/r/jailbreak/comments/458ao3/discussion_changing_time_date_settings_to_jan_1/ :)
[10:18:21] <kurteknikk> Question: I made a stupid mistake of converting a 90GB collection to capped (30GB) in mongodb 2.4, and the query has been running for over an hour with the mongo completely blocked. Is it safe to kill this operation ?
[10:56:28] <alezz> Hi all. Is there a way to $project the _id hex string value in a new field? I need it because I need to $lookup the _id with a foreign field which contains the hex value.
[10:57:45] <Derick> just the hex value, and not the objetid?
[10:57:50] <Derick> objectid*
[10:57:56] <Ceunincksken> Have you tried the aggregation framework?
[10:58:54] <Derick> Ceunincksken: that's what the $project refers too
[10:59:30] <Derick> alezz: I don't think there is an A/F operator that does that.
[11:01:21] <pchoo> I'm struggling with obtaining the data I want right now :| I have a collection with data like this: {name: 'whatever', states: [{type: 'a'}, {type: 'b'}]} and I want to find all records which have type:a, but NOT type:b in their statues array
[11:02:26] <Derick> pchoo: I think you should use A/F for that...
[11:03:18] <Derick> db.col.aggregate( [ { $match : { states.type : 'a' } }, { $match: { states.type: { $ne: 'b' } } } ] );
[11:03:38] <pchoo> Derick: Ah, ok, I was unwinding and trying to group on it
[11:03:46] <pchoo> but that just makes it much simpler haha, thanks!
[11:05:38] <pchoo> Derick: Hmm, would that just find other records which are $ne 'b', such as 'a'?
[11:07:14] <pchoo> oh, hmm
[11:07:30] <pchoo> that does get me a different result! Thanks! I'll look into it further :)
[11:21:25] <alezz> Yes Derick, it was something like "$project _id.str in my_id_str; $lookup on my_id_str = foreign_id", but I've already searched a lot and I felt the presentiment there's no operator that does that... Thanks for your reply anyway! :)
[11:22:23] <alezz> It would be usefule to have an operator like "$dateToString" also for object ids
[11:29:51] <Ange7> Someone can explain me why when i make "show collections" in mongo shell i have my list of collection but when i try with PHP driver $db->getCollectionNames() i have a empty array ? :/
[11:30:18] <Derick> alezz: yeah, but then again, you can argue that your lookup table should just have used an objectid...
[11:30:27] <Ange7> (my connection work cause if i try in PHP $collection->findOne() i have one document from my collection
[11:30:46] <Derick> Ange7: which MongoDB version, and which PHP driver version?
[11:31:39] <alezz> Derick you're right, but when you work on someone else's code... :D
[12:07:44] <lipiec> Hi, I have multiple MongoDB standalone instances. I would like to run some mongodb command on each of it at the same time. What is the best tool to use?
[12:58:36] <Ange7> Derick: mongo 3.2.1 / PHP Mongo 1.5.7-1
[13:00:10] <Ange7> Someone can explain me why when i make "show collections" in mongo shell i have my list of collection but when i try with PHP driver $db->getCollectionNames() i have a empty array ? :/
[13:02:53] <Ange7> (my connection work cause if i try in PHP $collection->findOne() i have one document from my collection
[13:15:09] <Derick> Ange7: you need 1.6 for the new listcollections command support
[13:16:03] <Derick> 1.5.7 is 1½ years old...
[13:16:25] <Derick> and does not support everything in MongoDB 3.0 and 3.2
[13:16:37] <Derick> for 3.2 support, you need 1.6.12 IIRC
[13:17:10] <Ange7> Thank you derick
[13:52:08] <alezz> Derick: today I'm very unlucked. I changed the model with a DBRef but... https://jira.mongodb.org/browse/SERVER-14466 shows that looking up through .$id is not currently supported... Yeah!
[13:52:11] <alezz> :D
[14:27:32] <Ange7> How can i get the max value in collection ? i have 1000 documents with date column, i want the max date
[14:35:20] <Ange7> ?
[14:58:36] <Ceunincksken> Ange7: Have you tried this: db.collection.find().sort({_id: -1}).limit(1)
[14:58:51] <Ceunincksken> instead of sort by '_id' sort on your date field.
[14:59:20] <Ceunincksken> You can do it with the aggregation framework aswell but that would be an overkill.
[14:59:53] <alezz> https://docs.mongodb.org/manual/reference/operator/aggregation-group/
[14:59:57] <alezz> :)
[15:53:14] <Ange7> is it possible to create an empty collection and define that future documents has index on column «x»
[15:53:28] <cheeser> just create the index on your collection
[15:54:50] <Ange7> :D
[15:54:52] <Ange7> Thank you =D
[16:00:53] <jami> Hi. I'm doing a stress test with ab against a simple node-mongoose-mongodb application. This application just make a single mongoose.connect with default options. After the stress test, all functionality with db logic inside hangs. After the test none of the model. ... .exec(func() { /* never reached */}) callbacks are reached. The application itself is running but mongoose does not work anymore
[16:01:24] <StephenLynx> i suggest you don't use mongoose.
[16:01:59] <jami> StephenLynx: yes i use mongoose
[16:02:39] <StephenLynx> try not using it.
[16:02:45] <StephenLynx> mongoose is awful
[16:03:21] <jami> StephenLynx: What should i use on node site?
[16:03:27] <jami> native db driver?
[16:03:48] <StephenLynx> mongodb
[16:03:59] <StephenLynx> this is the name of the native driver on npm
[16:04:15] <StephenLynx> http://mongodb.github.io/node-mongodb-native/2.1/api/ these are the docs
[16:06:49] <jami> i will give it a try
[16:10:19] <jami> StephenLynx: What would be the node-mongo-nativ
[16:10:28] <jami> behaviour on heavy load
[16:10:32] <StephenLynx> great
[16:10:47] <StephenLynx> its about 6x faster than mongoose
[16:11:09] <jami> StephenLynx: *hehe great answer no i meant if the load is too heavy and the connection pool is full
[16:11:50] <StephenLynx> don't know.
[16:12:24] <jami> StephenLynx: Would it block or does it gave a clearly error with 'Too many connections' or something
[16:13:08] <jami> because the mongoose behaviour is not acceptable. It just hang and didn't create another connection
[16:13:23] <StephenLynx> no idea, I never hit that problem.
[16:30:49] <Waheedi> So it seems the only way to get out of this, is to make my clients understand they have two primaries somehow :)
[16:37:42] <hugealt> Is it possible to have a new collection's documents split across shards while the balancer disabled?
[16:37:55] <hugealt> Providing I've set up the appropriate shard tag ranges
[16:38:12] <cheeser> you can manually migrate chunks
[16:38:25] <hugealt> Is manually migrating the only way possible?
[16:38:47] <cheeser> either you do it or the balancer does.
[16:39:25] <hugealt> Is it possible to run the balancer on an 'insert' only mode? Where it only runs on inserts, not on document updates?
[16:39:35] <cheeser> no
[16:41:07] <hugealt> Is it possible to have per-collection chunk sizes?
[16:43:19] <Waheedi> I must say my issue is a network related issue and has nothing to do with mongodb :)
[18:00:04] <Lachezar> Hey all. N008-ie here. If I have an Customer[name]->* Order [description] (*line) ->1 Product [description]. How do I find all orders that have description that contains some text or their customer's name contains the text, or the name of any of the products in the lines contain the text?
[18:03:04] <daidoji> Lachezar: what kind of notation is that?
[18:04:56] <Lachezar> daidoji: Notation? None. Just made it up db.customer, db.order, db.product . Order has DBRef to customer, and lines that have DBRef to product.
[18:05:11] <daidoji> Lachezar: oh mongo doesn't do joins
[18:05:22] <daidoji> so you'll have to join in javascript or the application
[18:05:38] <daidoji> normalization is the enemy and all that
[18:06:37] <Lachezar> daidoji: So basically I load all DB collections and do join/filter myself?
[18:07:11] <daidoji> Lachezar: yeah, its a philosophy of Mongo to explicitly not do joins in the database
[18:07:37] <daidoji> so you'd have to run three queries with the filters you've run, and then do more queries to get the other data you need from there
[18:07:46] <daidoji> to achieve the same thing as an RDBMS join
[18:08:25] <daidoji> Lachezar: Mongo documentation typically suggests denormalizing your data to achieve best results
[18:08:31] <daidoji> so you may want to do that if its feasible
[18:08:36] <Lachezar> daidoji: Well… That seems to be where I went to, but I was wondering if there aren't any server-side functions to handle that.
[18:09:20] <Lachezar> daidoji: denormalizing 1-* relations is a horror-show, and keeping denormalized data in sync is a nightmare.
[18:09:31] <daidoji> you're preaching the choir bro
[18:09:32] <Lachezar> daidoji: Been there, done that. No thanks.
[18:09:40] <daidoji> I just use Mongo and hang out here a lot
[18:10:01] <Lachezar> daidoji: no harm meant!
[18:10:16] <daidoji> good luck
[18:11:16] <Lachezar> daidoji: I'm currently doing it with multiple queries, but it seems really counter-productive to move big arrays of data from the server to the client, just to send them back again :)
[18:11:42] <StephenLynx> yeah
[18:11:50] <StephenLynx> usually when I have a case like taht
[18:11:58] <StephenLynx> I just create separate collections.
[18:12:09] <Lachezar> daidoji: I also find the { 'field.$id' : { $in : ['....', '....] } } does not work :(
[18:12:31] <StephenLynx> sure, fake relations come at a cost in mongo
[18:12:38] <daidoji> yeah :-(
[18:12:40] <StephenLynx> so I don't take that lightly
[18:12:49] <StephenLynx> but its a viable option, given the scenario.
[18:13:01] <Lachezar> So I'm not really missing some magic here. Am I doing it right?
[18:13:04] <daidoji> I mean if your data needs joins, I'd suggest moving to an RDBMS or Graph Database if you have a NoSQL application
[18:13:13] <StephenLynx> that
[18:13:24] <daidoji> if that is at all possible
[18:13:29] <daidoji> otherwise you're doing pretty much what I would do
[18:13:42] <daidoji> (slash do every day because my boss likes to denormalize our Mongo collection)
[18:13:48] <daidoji> collections rather
[18:14:01] <StephenLynx> i still much rather have fake relations than complex sub-arrays, though.
[18:14:05] <StephenLynx> any day of the week
[18:14:26] <StephenLynx> complex sub array can bone you hard
[18:14:26] <Lachezar> daidoji: Well… I'm a Java+JPA stooge, so it's relatively easy to move to PostgreSQL (say). But the customer 'heard', that MongoDB scales and is fault tolerant and all that stuff, and said we're using MongoDB.
[18:14:40] <StephenLynx> kek
[18:14:48] <StephenLynx> clients in charge of decisions
[18:14:51] <StephenLynx> classic
[18:14:56] <daidoji> haha
[18:15:10] <Lachezar> StephenLynx: My customer is not the end-client though. I'm sub-contracting :)
[18:15:17] <StephenLynx> still
[18:15:20] <daidoji> :-(
[18:15:37] <StephenLynx> clients should just hand requirements, not technical decisions.
[18:15:44] <daidoji> maybe look for a new job?
[18:15:46] <StephenLynx> unless the requirements include support for some technology.
[18:15:54] <daidoji> Mongo is good at a lot of things
[18:16:03] <daidoji> and a wrench is too
[18:16:04] <Lachezar> Anyway. It gives me a 'quantum of solstice' that I'm not missing big on 'The MongoDB Way' to do complex searches.
[18:16:09] <daidoji> but I wouldn't use a wrench to hammer a nail
[18:16:19] <daidoji> Lachezar: yeah I think most people run into that
[18:16:22] <StephenLynx> i would denormalize it.
[18:16:29] <StephenLynx> still better than anything sub-arrays will give you.
[18:16:33] <daidoji> other NoSQL solutions tend to offer some limited form of joins
[18:16:40] <daidoji> but Mongo doesn't support them philisophically
[18:16:51] <StephenLynx> specially for arrays of objects.
[18:16:57] <daidoji> Second StephenLynx's suggestion
[18:17:06] <StephenLynx> arrays of simple values are ok
[18:17:08] <daidoji> its usually the most reasonable way to do stuff
[18:17:24] <StephenLynx> but arrays of objects are not too pleasant to handle
[18:17:45] <StephenLynx> you CAN work with them, but IMO is not worth it
[18:17:49] <Lachezar> StephenLynx: there is no array of objects here. All 1-* relations go to specific collections.
[18:17:54] <StephenLynx> hm
[18:18:07] <StephenLynx> ok, what is the case? I didn't geti t completely
[18:19:43] <Lachezar> StephenLynx: I have to find all (a page of) all orders that: 1. Have description matching my search, or 2. Have customer whose name matches the search or 3. Have any lines that have products with names matching the search.
[18:20:26] <StephenLynx> what is your model?
[18:21:04] <Waheedi> joannac: got that double primary solved :) it was an iptables binat issue
[18:21:10] <Lachezar> If I understand correctly you suggest adding a 'search' field to the Order that will contain all the text that can bea searched?
[18:21:33] <StephenLynx> no
[18:21:58] <StephenLynx> I thought you had a complex array and suggested creating a new collection, but since that is not the case, I'm asking for your model.
[18:22:05] <StephenLynx> so I can understand how is your data organized
[18:22:16] <Lachezar> StephenLynx: You want an UML Class diagram?
[18:22:32] <StephenLynx> I think UML refers to software
[18:22:36] <StephenLynx> I want your database
[18:22:59] <Lachezar> StephenLynx: sorry, I lost you.
[18:23:15] <StephenLynx> your document on how your database is organized
[18:23:22] <Lachezar> StephenLynx: What 'model'? There is no 'schema' in Mongo right?
[18:23:25] <StephenLynx> no
[18:23:29] <StephenLynx> but you always follow a model.
[18:23:38] <StephenLynx> let me link you mine
[18:24:01] <StephenLynx> https://gitgud.io/LynxChan/LynxChan/blob/master/doc/Model.txt
[18:24:15] <StephenLynx> the first line is the name of the collection
[18:24:22] <StephenLynx> the others are the fields and a description of them
[18:24:27] <Lachezar> StephenLynx: Bah. I've nothing coming close to this.
[18:24:29] <StephenLynx> separated by -------
[18:24:31] <StephenLynx> so
[18:24:48] <StephenLynx> who knows how is the data organized?
[18:25:13] <Lachezar> StephenLynx: I'm a Java stooge. DB is something Spring Data maps.
[18:25:33] <StephenLynx> ok, and how devs can know how to store the data?
[18:25:57] <StephenLynx> and to retrieve it?
[18:26:01] <Lachezar> StephenLynx: Ugh? What? repository.save(new Customer("Name"))
[18:26:15] <StephenLynx> and know they know the fields to store?
[18:26:28] <Lachezar> StephenLynx: Customer google = repository.findByName("Google")
[18:26:40] <StephenLynx> that doesn't say the fields
[18:26:58] <Lachezar> StephenLynx: No. it doesn't isn't that the power of MongoDB? No schema?
[18:27:08] <StephenLynx> one thing is not enforcing a schema
[18:27:17] <StephenLynx> the other is using the database like a trashbin
[18:27:23] <StephenLynx> instead of a cabinet
[18:27:51] <StephenLynx> you seem to be just throwing stuff into this huge bin with no rhyme or reason
[18:28:14] <StephenLynx> you got bigger issues than w/e I was trying to help you.
[18:28:26] <StephenLynx> did you see the link I pasted?
[18:28:50] <StephenLynx> mongo doesn't enforce that and if I screw up, I can insert data outside of those rules
[18:29:00] <Lachezar> StephenLynx: OK. I understand. Yes, I saw the link. I've got nothing like that.
[18:29:07] <StephenLynx> welp
[18:29:11] <StephenLynx> sorry, but you are boned.
[18:29:34] <Lachezar> Yeah. I think I got that.
[19:05:54] <csepulvedab> Hello guys, how are you?
[19:06:00] <csepulvedab> some one know how set the ntoreturn value?
[19:06:47] <csepulvedab> i see always is 1000, and i hace some querys that need return more than 5000
[20:22:45] <kryo_> hi, i'm thinking of storing a json object in a mongodb collection
[20:23:06] <kryo_> only issue is that the structure of the json object might change
[20:23:16] <kryo_> so i cannot define a schema for it
[20:24:00] <kryo_> is it possible to store the json object directly as a collection?
[20:35:36] <starfly> kyro_: store a JSON document in a collection, structure can change later, but of course, it depends upon how you will use the data as to whether you will need to make all of the documents consistent later
[20:43:33] <Ben_1> is there a way to create a document in java without a key?
[20:51:03] <orriols> Hi. I'm using MongoDB 2.6. I need, using the aggregation framework, to project a field whose value is an array expressions. Is that possible?
[20:51:49] <orriols> the straightforward approach gives the error " failed: exception: disallowed field type Array in object expression"
[20:52:12] <orriols> is there a not-so-obvious approach to achieve this?
[20:56:42] <kryo_> ty starfly, seems like it's not as simple as you say since i'm using mongoose, so i've come up with a different solution instead
[20:57:36] <starfly> kyro_: OK, but nothing I said was incorrect
[20:58:09] <kryo_> yup. it works totally fine for raw mongodb
[20:58:36] <kryo_> just my case is more complicated
[23:50:57] <anidonde> I have a question re: mongo ruby driver. Is this the right place to ask?