Log file Viewer

[00:18:53] <jp-work> hey guys, can I make a shared cluster on one machine where each shard is an instance of a replica ser?

[00:18:55] <jp-work> *set

[00:19:13] <jp-work> I tried here but it says I can't add a shard server on localhost

[00:19:15] <jp-work> :(

[00:32:53] <jblack> sorry, I haven't gotten to sharding yet

[00:35:04] <Robbie> I'm new to sharding too, but could this possibly be your issue? ...

[00:35:05] <Robbie> "If you use either “localhost” or 127.0.0.1 as the hostname portion of any host identifier, for example as the host argument to addShard or the value to the --configdb run time option, then you must use “localhost” or 127.0.0.1 for all host settings for any MongoDB instances in the cluster. If you mix localhost addresses and remote host address, MongoDB will error."

[00:36:38] <jp-work> it works when the instances are stand alone servers

[00:36:53] <jp-work> but if you try to make them replicas it doesn't work

[00:39:30] <Robbie> have you seen this guide? http://docs.mongodb.org/manual/tutorial/convert-replica-set-to-replicated-shard-cluster/

[00:39:35] <Robbie> might be helpful

[00:44:06] <jp-work> Robbie: thanks man

[00:44:15] <jp-work> I think I tried that

[00:44:18] <jp-work> will try again

[00:44:19] <jp-work> :)

[00:44:37] <Robbie> no worries, sorry im not a mongodb guru (yet :)

[00:48:42] <jp-work> we'll get there :)

[00:49:14] <Glace> Hey.. does hashed shark key and sharding usually add more iops?

[00:50:58] <jp-work> I can't see why that would happen

[00:51:25] <Glace> hehe ok.. I am suspecting.. that i was just expecting better results :)

[00:51:37] <Glace> I have maybe 1.5 more hit/s than before

[00:51:46] <Glace> but was expecting at least 2 times

[00:52:05] <Glace> the load is well balanced between both shards though... but only got 1.5 improvments

[00:52:17] <Glace> so I was expecting maybe the additional hashed index

[00:53:05] <Glace> was causing this

[00:53:38] <Glace> still have a lot of write lock

[01:49:50] <kondi> What'd be the optimal pool size for mongodb?

[01:50:08] <kondi> The app receives monthly 50-60k visits (and 100-500 at any given time)

[01:50:11] <jblack> I almost said "depends on the size of the turds"

[01:51:17] <jblack> https://github.com/mongodb/node-mongodb-native/issues/529

[01:53:25] <kondi> heh, Alright jblack thanks for that link.

[01:55:45] <jblack> I parse that as "anything > 1 won't really do much more than increase your susceptibility to threading bugs""

[02:05:02] <sinisa> can some make any "harm" inside query params , after the "command" , like db.col.find({ anything harmful here}) .. or its not possible at all ?

[02:13:37] <jblack> I don't know how, if so.

[02:14:29] <sinisa> im doing something, web related.. so im mapping directly json to params .. wondering what can go wrong :)

[02:22:18] <jblack> Yeah, I figured as much. I'm under the impression that it should be safe, but I'm nowhere near experienced enough for you to trust me on that

[02:25:00] <sinisa> that was a question for experienced :)

[02:27:24] <jblack> I know, but at least you don't feel ignored now!

[02:27:37] <sinisa> :)

[04:00:42] <AlwaysBCoding> I'm new to mongodb, curious how I should model this association. I have a collection for 'nfl_teams' which is all 32 teams in the NFL. Then I have a collection for 'nfl_matchups' which represent all the matchups from a given season. Each 'nfl_matchup' has a 'home_team' and an 'away_team' which each will reference a team from the 'nfl_teams' collection. Since I'm new to mongo, not quite sure how I should go

[04:00:42] <AlwaysBCoding> about doing this. Any suggestions?

[04:01:51] <preaction_> that sounds like relational data to me

[04:02:20] <preaction_> you can use DBRefs if you want, but unless you're going to take advantage of the document structure, you can still use relational databases

[04:03:49] <AlwaysBCoding> I should mention that I'm using Rails / mongoid for this

[04:04:25] <preaction_> and? model your data and choose the database based on your data model

[04:07:41] <AlwaysBCoding> yeah for sure, I'm using mongo because I think that it will help me elsewhere in my data model. Just curious how to go about modeling unique relations like this, haven't been able to find any good resources yet

[04:09:17] <preaction_> DBRef is what mongo has for referential data, you can also simply use IDs, which is what I tend to do

[11:22:34] <pellis> hello