Log file Viewer
#mongodb logs for Tuesday the 2nd of September, 2014
[15:53:26] <adamcom> you can have only one shard, but each replica set is a shard, so while a shard can have 1, 2, 3…12 replica set members, they are all a single replica set
[15:54:08] <adamcom> not sure why you would want one shard but 2 replica sets in any case - with how MongoDB sharding works, the concept doesn't make much sense
[16:07:35] <krion> db.adminCommand( { listShards: 1 } ) is unknow on 2.4 for me, but the doc say so, weird...
[16:12:12] <adamcom> the mongod processes don't really know they are shards - the mongos is where you should run any shard aware commands
[16:13:33] <adamcom> sh.status() for example, will give you the shards listed (and other information)
[16:17:17] <adamcom> sh.status will work, because it is basically just a javascript wrapper for querying data out of the config database - it will technically work properly as long as you have an appropriate config database on the instance
[16:17:51] <TheRob_> Hi guys! I have one question about types that are being assigned by using the PHP MongoDB drivers batchInsert. Apparently they seem to be randomly variating between String and NumberLong. For example: Document 1 => "my_carrier" : NumberLong(129), "my_type" : NumberLong(2) | Document 2 => "my_carrier" : "126", "my_type" : "2". How can I prevent this and how can this happen?
[16:18:06] <adamcom> rs.remove() and rs.add() (and many others) are helpers - designed to make it less cumbersome to run commands
[16:18:34] <adamcom> as is db.adminCommand for that matter - it is a shortcut to avoid having to specify the admin database before running admin commands
[16:19:05] <adamcom> so not redundant exactly - they do the same thing, but they are there intentionally to make our lives a little easier
[16:20:26] <krion> adamcom: i'm hitting a either Unexpected token ILLEGAL or can't use 'local' database through mongos", "code" : 13644
[16:23:04] <adamcom> replica sets exist independently of sharding, when you run rs. commands you are administering the replica set - so you run the command there
[16:24:28] <adamcom> so you have a replica set with host1, host2, host3 - you run rs.remove() for host 2, your replica set is now host1, host3
[16:25:01] <adamcom> the replicas set is still there, still functioning, and still a shard (assuming it was before) - it just has one less member
[16:30:54] <krion> adamcom: the thing is by lack of knowledge we reinstall each node that act as a shard
[16:48:26] <SpeakerToMeat> Question, is there any tool or method you preffer or use at all when anotating/designing schemas for documment collections? My mind tells me to use something like xsd but that's rpobably an horrible choice. This will be human read anyhow.
[16:50:46] <krion> Tue Sep 2 18:47:32.629 [conn4476] clone failed{ note: "from execCommand", ok: 0.0, errmsg: "not master" }
[17:03:39] <SpeakerToMeat> Question, is there any tool or method you preffer or use at all when anotating/designing schemas for document collections? My mind tells me to use something like xsd but that's probably an horrible choice. This will be human read anyhow.
[17:06:38] <kali> SpeakerToMeat: some design through the ODM they're planning to use. other just use example documents with javascript comments
[17:52:10] <elfuego> lol, so the problem is, before I can use mongoimport I have to wait until mongod gets started, but mongoimport doesn’t have any context on when mongod is started
[17:55:21] <elfuego> cheeser I need a way to initialize mongo with data, and not having to worry about mongod latency on startup
[18:02:37] <elfuego> cheeser need to script it in jenkins, and its causing some problems like DBClientCursor::init call() failed
[18:02:56] <elfuego> here is a snippet of the scrip https://gist.github.com/horaceheaven/66d9f8f1ea3d6df3eb94
[18:16:42] <dberry> I just moved my sharded replica set to a new datacenter and I now see latency issues. Are there primary/secondary issues that would generate latency issues?
[19:21:05] <einw1ck> Im busy migrating my application to mongodb from sql, what is the mongo equivalent of: " SELECT * FROM books WHERE title LIKE '%search%' "
[20:23:00] <samed> hi, i've removed a secondary from replset recently, now i see many logs like "replset couldn't find a slave with id 2"? how can i deal with this?
[21:16:28] <oblio> i didnt delete any documents, i just trimmed out maybe 20% of the data in all documents
[21:21:18] <cheeser> i dunno about the average. i would imagine that's cached given how expensive that'd be to calculate.
[21:37:10] <SpeakerToMeat> I'll ask something here that I did somewhere else today and produced a slightly undesirable outcome. Is it very frowned uppon to use references/relations over embedding?
[21:38:01] <SpeakerToMeat> I seriously think the right solution for my schema idea is references (actually, it's using ancestor arrays). But I want to know the general take of the "community" (the few active people here willing to answer)
[22:19:19] <chrismclaughlin> is there a way to keep versions of the data in mongodb? I am looking for something similar to what git does for source code
[22:44:14] <SpeakerToMeat> Are collections hard partitions, like tables in rdbms, or are they softer? that is, is it complex if I work with references, to relate documents in different collections?
[22:46:11] <Boomtime> but there is some details under the bonnet.. like storage space, lock contention and permissions, that collections in the same database share
[22:52:34] <SpeakerToMeat> from what I read ObjectId is an object type in bson and not just represented as a string, and it's a type 1 uuid basically, right?
[23:02:58] <rh1n0-> i havent worked with mongodb in a while. Are there any good tools/clients that can help with refactoring raw mongodb queries?
[23:47:06] <styles> If I use the mongodb bulk operation but a record conflicts with an existing index or something & errors out, does the whole bulk operation fail?