Log file Viewer
#mongodb logs for Monday the 24th of March, 2014
[02:00:32] <pengwn> I would like some info on whether to use record references or record embedding on shard instances which will give better performance?
[06:12:04] <Brij-DBA> I am looking for mongodb setup guide on linux. Can somebody pls share me easy step to set it up?
[06:38:36] <Brij-DBA> I am new to mongodb, so just need easy steps if any one has those handy, to setup one mongo db env on linux.
[06:53:50] <mboman> Brij-DBA, have you checked the official manual? http://docs.mongodb.org/manual/administration/install-on-linux/
[08:37:04] <rAg3nix> hi , i am new to mongodb, i am trying to import mysql database using mongoimport. but there is a limitation of 16mb whereas my db size in json is 350 mb. can anyone please tell me how to do it ?
[08:44:28] <rAg3nix> i am trying to do that i guess, and its not happening for obvious reasons !! how do i do it ? i tried using gridfs !! it stores the whole file how do i run query in that?
[08:45:38] <Mothership> joannac, I want to commit my project to git with db files included, so when I pull it, I can automatically get the updated db files without the need to copy them to the main db path
[08:46:25] <Mothership> I know I can use bashscript, but I just thought if it's possible to add more db paths
[08:50:40] <joannac> rAg3nix: you don't? import your rows as documents or something. you're going to get terrible performance if you store your whole sql database as a mongodb gridfs file...
[08:52:52] <rAg3nix> joannac: no no , what i am trying to do is , i have the base database , and there are 3 tables that i need to migrate to mongodb on a master slave replica first table was done , second the one i am trying to do is 350 mb and is refusing to go as a single document. how do i resolve it or how do i do it ?
[08:57:47] <joannac> rAg3nix: you have a single table which is >350mb which you are trying to insert as a single mongodb document
[09:02:25] <joannac> I'm just confused why you exported your whole table as one single JSON document
[09:04:01] <joannac> Mothership: I still don't understand your use case. You want to clone your database without using mongodump/restore?
[09:13:05] <the8thbit> How would I do something like Foo.findOne( { fieldA: bar OR fieldB: bar }, function( err, foo ) { //code } ); ?
[09:28:45] <the8thbit> Nodex: Thanks. If I want to do a query from my client, what would be the easiest way to do this? Just set up a route for a .find() on the particular collection, and then use ajax to pass it a query?
[09:32:42] <amitprakash> Hi, given an _id, is it possible for me to look for a document across multiple collections matching _id
[09:37:17] <the8thbit> Nodex: I guess what I was asking is 'is this a reasonable approach', though I suppose you answered my question :)
[09:39:33] <Nodex> if you can't see the security hole in that method then you might be biting off more than you can chew
[09:48:10] <the8thbit> amtiprakash: Basically, something similar to a reddit profile, where I have a list of comments made by all users, and I need to list just the ones made by the user who owns the profile
[09:50:13] <the8thbit> I have authentication, so for private info, I could just make sure that the user is authenticated using the same username as they make the query for?
[09:52:03] <amitprakash> Nodex, so theres nothing short of cycling though multiple collections and looking for the _id
[09:52:25] <amitprakash> Nodex, also, can two documents share the same _id when they lie in different collections
[09:55:53] <the8thbit> amitprakash: We're talking about the same thing, yes? I set up a route. say /mongo/userComments that takes a string or two as a query. I then have a mongoose query server-side that takes the string I sent the route and does the actual querying
[11:06:18] <amitprakash> Nodex, https://www.idontplaydarts.com/2010/07/mongodb-is-vulnerable-to-sql-injection-in-php-at-least/
[11:25:47] <morfin> which is being calculated when i add new document to simplify search when i got lots of documents and need to search by that criteria
[11:26:52] <Nodex> there is no functions in Mongodb - not in the sense of things like lower(foo) == lower('Foo')
[11:47:51] <hjb> howdy. according to https://jira.mongodb.org/browse/SERVER-1625 mongo core server doesn't work on big-endian
[12:25:54] <hjb> Derick: well, ok. php is no option for us. i guess the java client should also work?
[12:37:18] <traplin> Meteor.users.find({"services.facebook.name":friendName}}).fetch();, with this statement how would i exclude any documents that the field "_id" equals "personID"
[12:38:29] <Derick> {"services.facebook.name":friendName, '_id' : { '$ne' : personID }} ought to do it
[12:41:23] <hjb> Derick: but for the "standard" command line client mongo there's little chance, right?
[13:07:18] <katspaugh> Hi! What is the recommended way to handle connections to MongoDB, when using mongodb-native driver?
[13:08:00] <katspaugh> Should I open a single connection and reuse it throughout my app, or open new connections to the DB each time I need to query or insert?
[13:08:57] <katspaugh> Right now, I’m reusing a single connection, never closing it. I experience hangings on bulk inserts and querying collections.
[13:21:18] <LoneSoldier728> am i calling populate wrong http://pastebin.com/S2QmZwPv it does not grab the results even though they are there
[14:28:12] <cfinley> Could anyone explain to me why the reduce method needs to output a value that can run through reduce again and output the same value?
[14:31:34] <kali> cfinley: it enables to reduce in several stages. so you can have a first reduce step on each shard of your sharded database, then one on a unique node to finish the work. it also allows to avoid reduce steps that would require an enormous values array, sometimes exceeding the memory
[15:05:07] <G1eb> Hello, is it possible to have an array with max length, meaning that when I push something new to it, it automatically deletes oldes item and adds the new one
[15:05:24] <kali> i think the behaviour is to reduce by chunk of 1000 docs, so yes, it can. you must not make assumptions here anyway
[15:05:46] <kali> cfinley: also, be aware the reduce may not be called if the values[] has just one element
[15:06:24] <cfinley> kali: I did notice that. Does that mean any statistics should be done in the finalize?
[15:06:52] <cfinley> kali: I'm writing a method that takes user events and generates user sessions for analytics.
[15:15:30] <Derick> FrancescoV: you can pass the path you want to the command line when you start mongodb
[15:19:28] <kali> Nodex: it's actually not that clear. they have officialy decided that the debris in the south indian ocean have to be it, but i don't think there is an actual physical proof yet :)
[16:13:18] <G1eb> kali, Im trying this $push $sort and $slice approach but keep getting: RangeError: Maximum call stack size exceeded =/
[16:22:37] <viktor_> How do I make a query on several _id's and get the result as an array with documents <in the same order> as my list of _id's? I've tried with $or and $in ...
[16:33:03] <riceo> hi everyone. quick sharing implementation question. are there any best practices around config server placement? I was thinking of running the config servers on my shard servers
[16:45:28] <viktor_> rkgarcia: hehe yes, it feels like this is something that many people will want to do, so I'm wondering if there's a recommended way of doing it. I dont see an obvious and nice way of doing it.
[16:57:34] <the8thbit> How would query for info in a member of an array? Also, is there any easy way to query for data in _every_ member of an array?
[17:35:57] <ShortWave> Is there a way to append to a array in a subdocument? I can't seem to get that to work right.
[17:38:44] <ShortWave> Ah, I'm in the middle of a discussion about how to handle this part of the data model, one second. The exact use case here is in some doubt, apparently.
[18:12:28] <faef> if I want a hidden secondary to do a fresh initial sync, I delete all the data in dbpath: does this include the local db?
[18:16:04] <faef> 1TB, it's annoying but it's the only way to get a compacted database without downtime
[18:16:43] <faef> we just deleted a bunch of indices on the primary, so we want to reclaim that disk space. choice is either repairdatabase directly on the secondary or just delete the data and resync
[18:51:45] <NaN> I'm creating a new collection from a nested document, $project worked but because I need to reduce the docs I used mapreduce
[18:52:04] <NaN> it worked, all it's ok, but I don't get how do I change that _id, value schema from the output
[18:53:53] <NaN> I thought about finalize, but it's the same, "value" gets the finalize results and "_id" the initial maped data
[18:55:17] <rkgarcia> NaN: http://stackoverflow.com/questions/8416262/how-to-change-the-structure-of-mongodbs-map-reduce-results
[18:57:27] <Honeyman> Hello. If I have 4 equal hosts, what is the best idea for a replica set configuration to utilize them fully?
[18:57:34] <Honeyman> The docs say if I have an even number (4) of replica set hosts, I should run the arbiter, but it shouldn't be executed on the same host as any of the rs hosts, so that calls for the fifth host...
[18:57:40] <Honeyman> Am I right that the best I can do is "1 primary, 2 regular secondaries, 1 non-voting secondary"?
[19:03:02] <G1eb> in the docs on http://docs.mongodb.org/manual/reference/operator/update/push/#up._S_push they use such a wierd notation for push with slice at the bottom
[19:56:56] <ShortWave> when I use geoNear, that's running a query using the 2d geospatial index that I've created on my collection, yes?
[20:01:25] <benth> does mongodb do anything special for sparse arrays? (i.e. arrays where most values are null or undefined)
[20:11:53] <ShortWave> So, better to use a dbCommand (like geoNear) vs. a straightup geospatial query?
[20:48:48] <kudos> is it a bad sign that I need to use a hint to make mongodb not sort super slowly?
[22:47:21] <alexi5> based on your experience what advantages does this type of database offer you over a relational database ?
[23:02:30] <LoneSoldier728> hey so I am getting back this from populate... http://pastebin.com/sigfqXKu and trying to call results.stories say is undefined... even results.stories[0] says cant get 0 of undefined ... and also, there should be two objects in there not one... what am i doing wrong ?
[23:03:31] <ShortWave> So just out of curiousity...why is it "ensureIndex"? is that just some functional naming with more meaning than normal?
[23:07:41] <ShortWave> let me know if that solves it...or not. I'll look deeper at it if you still have a problem