Log file Viewer
#mongodb logs for Friday the 1st of August, 2014
[06:55:18] <CodePulsar> : error: ‘getScopedDbConnection’ is not a member of ‘mongo::ScopedDbConnection’
[06:58:21] <CodePulsar> Where the heck is the definition of ScopedDbConnection https://github.com/mongodb/mongo-cxx-driver/search?q=ScopedDbConnection&ref=cmdform ?
[07:13:38] <jianing> i have a question about sharding cluster. Does the chunk really store the document in a sequential disk space.
[07:39:56] <aydintd> Hello all guys, i have 2 replica shards contains 1 primary and 2 secondary mongos in it. In one of them i've about 270GB data and trying to keep my replication up but my replicas' always going to break. There was a misconfiguration on primary mongo about opLog size (its only 700MB) but i can't change it because of production. But i want to do a health check more efficient, what u guys suggest?
[07:45:53] <rspijker> Finally, what does “But I want to do a health check more efficient” mean? Replication isn’t for health checks, nor do I see anything else in your question which refers to efficiency or the lack thereof.
[07:46:29] <aydintd> i'm newbie on mongodb actually. I can't change it because new datas coming continously and i dont have any backups because of broken replications. I can't restart primary node
[07:47:22] <aydintd> my primary's oplog was set 700mb before me, i changed my replicas oplogsize 12gb but it didnt wokr out.
[07:50:55] <aydintd> yeah you're right but it's not my company's mongo, they're belongs to one company who i work for
[07:52:24] <rspijker> what causes the situation is probably that in the 2 days it needs to catch up, more than 700MB worth of oplog traffic is recorded on the primary
[07:53:08] <aydintd> i also run nagios on mongo replicas and getting service check timed out notifications on Mongo Memory Usage
[07:56:06] <rspijker> nagios checks timeing out could have a bunch of reasons, config issues, host overloaded, network down, etc.
[07:57:31] <aydintd> i also see that, when my primary oplogDate shows today, my replicas showing 2 days ago when they broke.
[08:07:17] <CodePulsar> wtf is going on here: http://paste.kde.org/pe7ajilcx/dpkeer ? at main:1468 I have scoped_ptr<ScopedDbConnection> dbConnection(new ScopedDbConnection("localhost"))
[08:21:58] <aydintd> it's says 17 hours, is it the needed time to cover opLog when it's set to 700MB isn't it?
[10:37:11] <hyperboreean> Hey, can I use $pull with $set in the same update (on different keys, not the same ones) ?
[12:20:25] <remonvv> kali: As you should, perhaps you should go stand in a corner and reflect on your laziness.
[12:41:07] <asteve> can you store historically relational models like user in mongo; follow-up question: is that a smart idea?
[12:52:34] <remonvv> asteve: Can you be a bit more specific. You mean moving a relational user database to MongoDB?
[12:53:31] <asteve> well, I'm fairly with other nosql's and I've always historically put a "user" model in an rdbms; I'm working on a new project and I would like to try out mongo
[12:54:35] <asteve> or rephrase it to something that needs two way lookup, both on some key like email address and on id
[12:54:50] <kali> asteve: i know SQL database quite well, for having work with many of them for the best (or the worst) of 10 years. today there is no task or use case i've encountered for which i would not consider using mongodb first
[13:07:55] <asteve> if I create a document that is {name: "asteve", email: "asteve@asteve.asteve", password: "cleartext"}, can I find the record by any of those fields?
[13:10:59] <kali> asteve: you need indexes on fields or fields combination you'll use for querying. just like in SQL
[13:21:52] <asteve> what is the purpose of the "1" in this example: db.people.ensureIndex( { "address.zipcode": 1 } )?
[13:36:23] <remonvv> asteve: It makes it valid JSON in that specific example, for compound indexes (e.g. {a:1, b:-1}) 1 or -1 can be used to specify the order relative to the earlier index field.
[13:37:59] <remonvv> There are a number of operators that take any value without that value having any real significance
[14:59:22] <cozby> hi, I'm inserting a lot of records from a nodejs app to mongo and after exactly 60 seconds I start getting timeout errors. Is there a config setting for this on the mongo side that I should be aware of ? Any help would be appreciated, I'm at my wits end on this.
[15:10:02] <remonvv> Can you show us the specific error? Inserting a lot of documents should not result in a timeout.
[15:13:18] <mikebronner> looking for a way to combine multiple aggregate queries into one resultset, so I don't have to make multiple calls to the database (to reduce overhead): http://stackoverflow.com/questions/25040792/return-limited-number-of-records-of-a-certain-type-but-unlimited-number-of-othe
[17:50:54] <doghugger> hey everyone, I'm having some issues with the ruby mongo driver. does anyone know if it supports exist? I've looked at http://docs.mongodb.org/manual/reference/operator/query/exists/ but not been able to implement it in ruby
[18:19:51] <defunctzombie> When using the $match piece of the aggregation pipeline, how would I specify that i want to match on documents with a created_at field less than some value?
[19:25:07] <kali> mozzie: oué, mais ca n'existe pas pour mongo. tu peux materialiser une liaison equivalente à une FK, mais mongo ne la validera pas (en verifiant que la destination existe) et ne te permettra pas de faire des requetes portant sur les deux collectiosn ene meme temps
[19:33:43] <kali> bha dans ce cas, il te faut organiser la donnée pour que cela n'arrive pas, par exemple en embarquant une table dans une autre
[19:38:00] <mozzie> merci pour les info, je vais aller regarder un peu la doc et voir si c'est vraiment ce dont j'ai besoin pour mon projet.
[19:59:37] <rickitan> Guys does anyone know if there is an open bug for using the $addToSet operator after doing the query with $in ?
[19:59:59] <rickitan> Ex: Topic.update({_id: {$in: topicIdsToAddOrganization}}, {$addToSet: { organizations: organizationId }}
[20:03:20] <Es0teric> i have 13k items from a json array that need to be split to allow for quicker insertion
[20:33:37] <Es0teric> i might do additional processing later like adding extra keys/values to the arrays
[20:51:55] <Es0teric> first i have moloquent, which i call on the database seeder file i created… with guzzle, guzzle is a HTTP package that allows me to get api data… i then use that data to insert it into mongo using moloquent
[23:00:18] <klevison> for gelocation application based, mongoDB is the best choice? (5 millions requests per day)/600GB DATA (SQL Server)
[23:48:22] <Kettlecooked> Looking at Trello as an example, their unique ID's might look something like 'q7nkmEqZ' - much tidier than a MongoDB ObjectID. Where would I start if I'd like to incorporate unique ID's that look more like Trello style?
[23:50:29] <Kettlecooked> I suppose I just make the _id type a String, and for every new document I generate an ID and check that it's not already available? If available - generate new - repeat until satisfied?