Log file Viewer
#mongodb logs for Thursday the 13th of August, 2015
[06:16:36] <Boomtime> aps: the download of ops manager is for 'evaluation' only and wants contact info to talk to a sales rep, so no it's a paid product
[06:18:03] <mbuf> when I restore a replica set into a new node, I still see the old members when I do rs.conf(), even though when I start the new instance with a different replica set name in /etc/mongod.conf. How can I remove the old members with force?
[06:24:41] <joannac> mbuf: stop the node, remove local.* from the dbpath, and then start and initiate
[10:43:55] <NoReflex> hello! I have some slow queries so I added a new index to make them faster. Up until now the query used a large index (1.2 GB) so I added a new sparse index (47 MB).
[10:44:36] <NoReflex> I can contruct my query so that it uses one index or another but the query runs just as slow
[10:45:36] <NoReflex> so in the log I can see planSummary: IXSCAN { xti: 1.0, xrt: 1.0 } - this is the new index
[10:52:26] <NoReflex> I'm using MongoDB 3.0.5 on Arch, WT with snappy on a notebook with i7 3517U CPU, 10 GB RAM, HDD 750 GB and the database size reported by show dbs is 10 GB
[14:21:59] <xcesariox> cheeser: line 47 seems fine. but why line 17 shows an error that i am not defining it?
[14:38:19] <xcesariox> cheeser: okay i added the function but now it prompts this error https://eval.in/416381
[14:39:04] <hypfer> hello, I have a simple aggregate group query and I'd like to multiply the average value by 8760 and then divide it by 1000. here's the current query: http://paste.debian.net/291966/
[14:40:04] <cheeser> xcesariox: i'm pretty sure you can't define a function as a function declaration parameter. you can *invoke* a function that that, though.
[14:50:55] <xcesariox> cheeser: i copied the whole function so that you know what i was trying to do.
[15:11:06] <bogen-work> I'm trying to use amid-rest. I'm able to query and get. When I post a new document is created, but the supplied json string, while not reject by the request is ignored. amid-rest returns the location (an empty document apart from the assigned _id) and waits for the next request.
[15:34:40] <bogen-work> ok, got POST's working.... (I had application-json rather than application/json for Content-Type)
[17:26:20] <gcfhvjbkn> i've created a server-side function by populating the system.js table as described here
[17:27:03] <gcfhvjbkn> cheeser: i've semi-solved that one, but yeah i'd love to know of a standard way to do that
[17:27:40] <gcfhvjbkn> now i can access the function in the mongo shell after invoking db.loadServerScripts();
[17:28:13] <gcfhvjbkn> i do this: new BasicDBObject("_id", new Code("getNextSequence('userid')")).append("rawUserId", rawUserId)
[17:28:26] <gcfhvjbkn> and it seems to insert "getNextSequence('userid')" as a code object without invoking it
[17:31:03] <gcfhvjbkn> this is my take, which doesn't work to well because i have to "case" every json type manually and even then you need to serialize Arrays and Documents in a special manner
[17:31:36] <gcfhvjbkn> you'd think the api will provide a method that builds the whole document tree for you
[17:37:35] <lufi> hi, is there something wrong with my query? http://pastebin.com/dybHbT8Z .I am getting an "invalid operator $dateToString" error
[17:56:22] <deathanchor> any know of code/query to find dups so I can build a unique compound index?
[18:02:35] <deathanchor> found something http://blog.mongolab.com/2014/03/finding-duplicate-keys-with-the-mongodb-aggregation-framework/
[18:51:54] <lucha> Is there a way to generate a many to many rellationship between two collections in mongoose?
[19:17:03] <Pinkamena_D> I am trying to store a field in a document which contains a type of mapping from filenames to file hashes. The majority of queries will be giving the file name of one file and wanting back a mapping of all filenames to file hashes in that document. Because I can not use filenames as the keys directly, what is the best way to store this data?
[19:17:47] <Pinkamena_D> I can think of many possibilities, such as using an array of documents, or using two parallel arrays.
[19:41:16] <gcfhvjbkn> i tried bson.types.Code and bson.BsonJavaScript, but they do a different thing:
[19:41:55] <gcfhvjbkn> the code i sent gets evaluated when i type db.users.find() in mongo, not at the moment when the data is saved
[19:42:31] <gcfhvjbkn> i suppose it might not be one of the bson.types._ classes because in this use case nothing needs to be stored on disk
[19:44:57] <cheeser> i don't think you can via java. executing serverside javascript isn't a big priority for the java driver.
[19:46:28] <gcfhvjbkn> can't really rewrite it in java, i want to rely on it to be run on the server-side
[19:53:05] <happiness_color> getNextSequence("userid") { collection.findAndModify(query, new BasicDBObject("$inc", new BaseciDBObject("_id", 1));
[19:54:43] <happiness_color> http://docs.mongodb.org/manual/tutorial/create-an-auto-incrementing-field/
[19:56:56] <gcfhvjbkn> what happens if two clients try to increment the same key simultaneously? i wager one will get an exception on insert
[19:58:44] <happiness_color> if you say findAndModify and do $inc for the field, that should be atomic
[21:02:13] <happiness_color> keep another (indexed) field with concatenated values?.. :/ nothing better comes to mind
[21:44:41] <Petazz> So I just set mongoengine to use replicasets by setting a keyword arg replicaSet="rs0", so far so good. Then I tried going into the primary and calling rs.stepDown() so a new primary was elected
[21:44:56] <Petazz> Then it kept giving me pymongo.errors:AutoReconnect: not master and slaveOk=false