Log file Viewer
#mongodb logs for Monday the 29th of June, 2015
[00:33:58] <jecran> db.test.find( { a: { $exists: true } } ... what is the most efficient way to search for 2 different values? (can be from different records)
[10:20:24] <leporello> Hi. I'm using mongoose and get an error: { name: 'MongoError', message: 'getaddrinfo ENOTFOUND' }
[10:27:56] <leporello> Hm. Missing the parenthesis in ternary operation caused wrong string to connect.
[11:16:50] <flok420> I'm trying to count the number of occurences of a certain value. e.g. 0.1...0.2 should be counted, 0.2...0.3, etc. I tried the following but it says that it did not expect ':' without any further indications which : it did not expect. any suggestions?
[11:16:50] <flok420> db.data.group(key : { }, cond : {'ts' : { '$gt' : 1435487780 } }, initial : { counts : [] }, reduce : function(curr, result) { result.counts[int((curr.value - 0.0) / (0.24-0.0) * 10.0)]++ })
[11:29:57] <symbol> If I'm only doing bulk writes, is there and advantage to using the Bulk api vs inserting with an array?
[11:33:44] <symbol> Good point, I'm working on some lesson plans but am using the MongoDB shell to teach it. I'm debating going over bulk operations due to the variety of driver handling.
[11:39:39] <zamliman94> the geoNear sorting is lost with group, and i cannot use geoNear after group, because geoNear has to be the first in pipeline order
[16:46:46] <dorongutman> I have a social activity-like collection, with documents that look like this: http://pastebin.com/kq6qnUtJ. as you can see there’s 3 “status” creations and then a “delete” on one of them.
[16:47:28] <dorongutman> how can I get all the items that were created but not deleted (if they were deleted I need to get just the delete) ?
[16:48:16] <dorongutman> I tried with the aggregation framework but I can’t seem to get it right - $group loses the data on the document, and I don’t know how I get the delete itself (which must be the most recent one)
[16:51:20] <GothAlice> dorongutman: $group on the "group" dbref, collect ($addToSet) the actions to create a new list of all actions for each group, then $match on that ($and the "actions" list containing the action you are looking for and $not the delete action). You can then project/unwind/etc. from there.
[16:52:38] <GothAlice> However, likely if it exists it was created (this is a simplified case), so you can very simply $group on the "group" field, collect the actions ($addToSet), and $not the actions containing "delete".
[17:18:43] <dorongutman> GothAlice: thanks. I’ve gotten up to the collect the actions with $addToSet (I think) but now how do I $not the actions contatining delete, and how do I get the original document (instead of just the grouped properties) ?
[17:18:50] <dorongutman> this is what I have: db.EventLog.aggregate([{$match: {'group.$id': ObjectId('5591566e279871e5267b23c6')}}, {$sort: {'loggedTimestamp': 1}}, {$group: {_id: {"effectedId": "$effectedId"}, "actions": {$addToSet: "$action"}}}])
[17:20:56] <dorongutman> GothAlice: actually - I do need the “delete” actions themselves, but not the “create” on the items that got deleted
[18:28:34] <fxmulder> if I have a replica set that used to have sharding enabled but now for some reason doesn't think it is enabled, how do I get it set back up as a shard member?
[19:10:37] <grandy> hello, wondering if there is a recommended tutorial for understanding what mongoose can do .. i've used "raw" mongo before but now working on a codebase that uses mongoose, trying to figure out which approach will be best for writing some report
[19:11:20] <StephenLynx> most people here just advice to not use it. but since the codebase already uses it, I don't think thats an option.
[19:16:59] <StephenLynx> node, an now io, is one of most performatic, if not the most, scripted runtime environment for web.
[19:18:33] <grandy> yeah hoping es6 + strong mode and strict mode will hep prevent some of that oddness... not minding es6
[19:20:05] <StephenLynx> that is why people use in lint a settings that force the use of === and !==
[19:20:36] <GothAlice> So "ignore half your language, since it's crap" is acceptable. I find that weird.
[19:22:12] <GothAlice> StephenLynx: Which any language, esp. ones supporting coroutines, can do. ;^P
[19:27:59] <GothAlice> I mean, if you have to pick a virtual machine to run bytecode in, it's hard to beat the JVM. Really.
[19:32:11] <GothAlice> Hint: if it's "interpreted", it's running a bytecode virtual machine behind the scenes. The compilation step is just hidden from you in most cases. (I.e. the typical "PHP throwing out the compiled bytecode after each request" problem solved by bytecode caches, etc.)
[19:32:40] <StephenLynx> I know, but woudn't using JVM require it to be compiled to bytecode beforehand?
[19:32:50] <StephenLynx> "{"name":"Anonymous","signedRole":"Board owner","email":null,"id":"6dfa4a","subject":null,"markdown":"a","message":"a","creation":"2015-06-29T19:05:46.467Z","files":[{"originalName":"397.jpg","path":"/second/media/10.jpg","thumb":"/second/media/t_10.jpg","name":"10.jpg","size":64488,"width":447,"height":400}],"postId":51}"
[19:33:01] <GothAlice> That VM may boil things down to machine code in the end (JIT and whatnot), but it's still using a parser -> AST -> compiler -> bytecode translation process.
[19:33:42] <GothAlice> StephenLynx: Speed isn't everything. The JVM rocks because of speed, AND everything else it does.
[19:36:09] <webjay> would it the be possible to have a replica that some can only connect to, so their queries can't hurt the master?
[19:49:47] <cheeser> replica sets aren't really the best option for scaling reads. you'll want sharding for that.
[20:51:30] <hahuang65> joannac: thanks for your help in our ticket a few weeks ago :) Never got a chance to respond.
[21:44:21] <kaliya> Hello, I have a dedicate partition for mongo data. With a replicaset on mydb, what's the best way to not make mydb.* files to fill 100% the partition? Is there some limit option? Mongo 2.4
[22:23:24] <benjick> Is there a way to do this I'm trying to achive? https://gist.github.com/benjick/2a262acc33bf227c038c
[22:47:48] <grandy> can anyone explain the meaning of the terms in an aggregate group clause? not sure i understand the quoted string with a $
[22:50:50] <Doyle> Hey. How much are the Enterprise Advanced subscriptions? Just looking to get OpsManager for backups.
[22:57:08] <grandy> newb question: for this query why does it fail if I substitude _id with something else? db.mycollection.aggregate([ {$group : {_id: "$a", count: {$sum: 1}}}])
[23:00:22] <grandy> any recommended way to quickly get up to speed on aggregates? any good online tutorials?
[23:34:46] <Freman> [[09:27:21.202]] [ERROR] { [MongoError: Executor error: Overflow sort stage buffered data usage of 35067470 bytes exceeds internal limit of 33554432 bytes] name: 'MongoError' }
[23:35:51] <Freman> {task_id: this.task._id}, '_id date duration pid code server failed', {sort: {date: -1 }, limit: limit}
[23:50:25] <Freman> I suspect it's loading up the 10+ mb 'messages' field for sort/limit even tho I've not asked for it