Log file Viewer

[00:29:34] <Moussekateer> anyone have experience with pymongo?

[00:29:47] <Moussekateer> I'm trying to optimize this query edits_collection.find({'user_id': user['_id'], 'timestamp': {'$gte': start, '$lt': end}}, fields=['timestamp', 'page_id'])

[00:30:27] <Moussekateer> I've created an index on {user_id: 1, timestamp: 1} but I'm not sure if it's being used

[01:08:42] <capnkooc> hi, is posible to get the date stored on a objectid with $project and the aggregation framework?

[04:18:55] <domo> let's say i do a mass insert of data one day and create indexes on a few keys

[04:19:02] <domo> a few days later, i insert a lot more data

[04:19:09] <domo> do I call ensureIndex again or reindex?

[04:19:17] <domo> or neither

[04:28:25] <X-Jester> domo: i think your indexes are created on insert

[04:28:37] <X-Jester> domo: so unless you want to change or create a new index, you're fine

[04:31:14] <Glacee> any recommendations to tune oplogs? queries to oplog are around 800ms

[04:50:59] <domo> X-Jester: ok thanks

[10:53:27] <jasper^off> hej guys, how does the mongodb _id work...it's generated client side?

[10:57:41] <Nodex> yes

[10:58:06] <Nodex> well sort of

[10:58:16] <Nodex> you cna get one in your driver is what I mean

[13:43:31] <shmoon> hi anyone on?

[14:16:25] <saby> I have a collection containing subfields like this http://pastebin.com/BjbHfgku

[14:17:05] <saby> i want to make a query such that I can fetch records containing all the 4 status fields but the fields should have either a processed value or failed value

[14:19:21] <saby> something like (OR(status.sentiment=processed, status.sentiment=failure)) AND (OR(status.aaa=processed, status.aaa=failed)) AND ...

[14:25:46] <pats_Toms> mongodb have something like lists on redis?

[14:27:49] <Derick> mongodb has arrays for field values... Not quite sure what you mean by a list?

[14:28:47] <pats_Toms> seems like I have some wrong logic in my head

[14:44:14] <saby> anyone?

[15:35:03] <pellis> does mongodb need index on a sort field when querying for a sorted collection?

[17:23:23] <skrammy> hi all. im wondering.. what would happen if i stopped a --repair in progress? really bad things?

[17:24:14] <penzur> chaos

[17:25:23] <skrammy> bummer

[17:25:40] <skrammy> worse than cutting power to an instance in non-repair mode?

[17:26:24] <skrammy> basically, im repairing a 5million object collection and was hoping itd go faster so that i can go out in 3 hours.. not clear if itll be done by then and then there are other large collections itll have to repair too

[17:28:06] <skrammy> i wish i could stop it and change the EC2 instance to bigger...

[20:14:12] <lobbin> Any hints on how to store localized content for easy access? Should I mimic the complete document or should I create attributes for each language, such as name, name_sv, name_dk, etc?

[20:27:46] <AAA_awright> lobbin: Generally different encodings with different media types, languages, etc, get different resources, so I'm inclined to say put them in different documents with a "language" tag

[20:31:57] <lobbin> AAA_awright: Ok, in this case it'll only be a few fields, so perhaps that would be overdoing it, still thinking about it :)