Log file Viewer
#mongodb logs for Wednesday the 4th of June, 2014
[00:26:41] <jason0_> proteneer_osx: will most drivers distribute queries over multiple replicasets? For example a complex find operation that has to scan all elements, will this become distributed with replicasets?
[00:35:28] <jason0_> proteeneer_osx: i see. but it would be up to the driver to distribute the query.
[02:38:00] <gancl> og01: http://stackoverflow.com/questions/24017372/nodejs-mongodb-findandmodify-exception-e11000-duplicate-key-error-index Thanks!
[04:50:12] <n4l> anyone know why I would be unable to drop a db? in the mongo shell im doing use [database];
[04:53:28] <n4l> [ { _id: 538ea1d644f4078d5f000001, name: 'Arthur Dent' }, { _id: 538ea1d644f4078d5f000002, name: 'Ford Prefect' } ]
[07:18:24] <kali> Egoist: if the config servers are fine, queries on magic collections from the "config" database will work
[07:20:46] <rspijker> Egoist: general tip, just state your problem instead of saying you havea problem and waiting for someone to respond to that. If you state your problem and someone knows about it, they’ll probably respond
[07:30:17] <Egoist> this is what i have in my mongos log: could not verify that config servers are in sync
[08:25:15] <Egoist> is there any way to check configsvr from mongos instance, before i connect them to each other?
[08:30:12] <joannac> point: {"key" "Master..."} does not macth at there's another field (name: test1)
[08:30:39] <Egoist> i want to check that's configsvr is waiting for connection, but i want to do this from another host
[08:32:40] <josias> @joannac but how? i the software doesn't know the name... why it isn't possible to make a deep elemMatch?
[08:48:27] <abbood1515> http://stackoverflow.com/questions/24032571/how-to-search-in-mongodb-by-querying-multiple-fields-in-the-same-document
[09:26:05] <AlexejK> when trying to see if Mongo needs more RAM under e.g Aggregations, is looking at serverStatus.recordStats.accessNotInMemory a proper thing to do?
[09:31:53] <rspijker> if you are under continuous load and you see a lot of paging in/out, then you would want to increase RAM
[09:56:09] <AlexejK> so is my understanding correct that high page fault is bad? (sorry, not so good with that)
[10:09:39] <rspijker> a page fault basically means that mongo accesses a piece of virtual memory that is not in RAM. Which in turn means that it has to go and retrieve the info from disk and load it into RAM
[10:09:39] <AlexejK> then i think i have a problem as during the heavy operation time (the one i've seen) page faults on primary suddenly spiked to 100-150
[10:10:15] <rspijker> if you are accessing information that’s old or not used often, it’s always going to have to be paged in
[10:12:55] <rspijker> which means that during that minute, it does an average of 100-150 page faults per second
[10:13:59] <AlexejK> yes, we have some aggregation jobs running.. they use both Map Reduce and Aggregation framework operations
[10:15:10] <AlexejK> we have a java app that starts behaving relly sluggish and DB queries suddenly return nothing when looking up a value that has just been written to DB (at least sent to socket) and this gives errors back through our API. this only happens when these aggregations are running
[10:15:53] <AlexejK> this specific entity is read from primary to ensure we respond fastest possible back and don't wait for replication to slaves
[10:17:27] <AlexejK> basically we create a session key and this key lookup fails for us when mongo is running a big operation.. this also happens if we do a delete operation on many entries.. which partially is udnerstandable but currently my guess is that IO performance is a problem and possibly memery (not ruling out app code problem as well)
[12:34:57] <minimoo1> i have a schema containing array of objectids, i was wondering if there is a way to remove one of the values in the array with a single query
[12:42:30] <AlexejK> humm suddenly mms-monitoring agent started complaining about "Either the server detected the possibility of another monitoring agent running, or no Hosts are configured on the MMS group." And all i did was upgrade from agent 2.1 to 2.2
[12:43:56] <AlexejK> dont have multiple agents running in that installation (however do have several different agents for my group for different server groups, but they are isolated from eachother)
[12:49:01] <minimoo1> i found this http://stackoverflow.com/questions/19786075/mongoose-deleting-pull-a-document-within-an-array-does-not-work-with-objectid
[13:41:50] <minimoo1> can someone tell me please why this doesn't work? http://jsfiddle.net/gD8ft/ (ignore syntax errors if missed anything)
[13:49:51] <kali> minimoo1: time to go to the mongodb shell and try to make it work without mongoose
[13:50:37] <kali> minimoo1: $pull: { appointments : { '_id' : mongoose.Types.ObjectId('538f04e45a4e06a01b000013') } } }. you want $pull: { appointments : mongoose.Types.ObjectId('538f04e45a4e06a01b000013') } }
[14:08:11] <jaggerwang> I installed mms and mms-backup, there is a replica set being monitored and backuped, all is going well. But when I terminate the backup, it keeps status in teminating, and when I try to remove hosts from monitoring, it says “This host cannot be deleted because it is enabled for backup”.
[14:37:59] <AlexejK> out of curiocity, is there a reason why MMS does not allow logical grouping of server+agent combos? I mean we have several separate/isolated datacenters and want to manage them in one place.. But not it seems (i think this changed recently) we have to have 1 group per isolated datacenter
[15:17:45] <Tulsene> I begin with mongoDB and trying to perform authentification on standalone cluster and after activate auth in /etc/mongodb.conf I'm trying this command db.createUser( { user: "siteUserAdmin", pwd: "password", roles: [ { role: "userAdminAnyDatabase", db: "admin" } ] } )
[15:18:59] <Tulsene> and if I try addUser instead of cresteUser I got this : Roles must be non-empty string
[16:03:50] <kesroesw1yth> How can I query for documents where 1 or more objects don't have a key by a particular name, or where that key exists but is empty?
[16:05:21] <AlexejK> there is also a few"see-also" links that i think you can look into.. I think you can do it in one query with logical or
[16:05:36] <kesroesw1yth> The issue is that the documents have any number of properties which contain objects, and I need to check all of them recursively and return true if any are missing a given key.
[16:06:48] <kesroesw1yth> Nodex, I can do that, but the recursive part is what I'm having trouble with.
[16:07:38] <AlexejK> I don't think you can check EVERY key in the document to look for a sub-key existance, can you?
[16:07:40] <kesroesw1yth> Basically return any document where any nested object it contains does not have a property by a name I provide in the query.
[16:08:24] <kesroesw1yth> Nodex, I do know that much, but the issue there is that I can't know the subdocument names beforehand.
[16:09:22] <kesroesw1yth> I tried that. Perl is a real pain about nested hashes when they come from Mongo for some reason.
[16:09:40] <Nodex> if you're looking for any key that doens't exist or is null then you will have to loop EVERY key in a document and check it
[16:12:14] <kesroesw1yth> Not sure why but I can't seem to iterate through nested hashes like usual if the original hash came from Mongo.
[16:13:52] <kesroesw1yth> I can create a hash when retrieving the object from Mongo, but I can't seem to refer to its keys with foreach(keys%$hash) like normal.
[16:24:12] <AlexejK> In MMS, when i look at OPCounters, I think i understand the "getmore" graph, but what is counted in "command"
[16:29:35] <q851> Question for you ladies and gents: I have a host running mongos that connects to my cluster. Can I update the binary while mongos is still running THEN restart the process? I need to minimize downtime while upgrading the 2.0.X mongos components.
[17:40:59] <subway> Howdy... I'm in the process of taking a snapshot of a couple exsiting databases in one replica set, and seeding a fresh replicaset with them.
[17:41:45] <subway> With mongo stopped, can I safely just grab all but local.* and journal from my source data dir, and drop them into an empty data dir on the first node in my new replica set?
[18:08:38] <ddod> Mongo noob here using node native: I currently have a function that makes sure people can't use the same value of something that someone else has (e.g. to make sure every user has a unique email address). Is this possible to do in the findandmodify function by itself? I don't really understand the ensureIndex function and when I would call that.
[18:21:47] <tscanausa> so you would create a unique email address on email address and then try to insert an error should raise when you do that.
[18:32:00] <gancl> Hi! Mongoose How to findOneAndUpdate insert new score when not exist and increase score when exist?
[19:16:46] <Dieterbe> hey, in my config i have 'noprealloc = true' and also 'journal = false', but yet when i start mongo it always tries to allocate 3GB of journal files ><
[19:40:36] <ranman> Dieterbe: the only difference in my config is that I have a space after the config
[19:47:48] <unkleara> Hi, I have a existing rails app using mongodb, I am building a feature with node and need to query the same database. Are their any resources to guide me how to set it up properly? Do i create a new db user for the node app or use same as rails? I tried using mongoose but I think it requires a scheme defined and I dont want it to create new collections. I just need to read from the existing database…
[20:05:03] <mistawright> hi guys i need some help. my server ran out of space due to me filling up my mongodb databases. how can i clear these databases? I cannot seem to login into mongo on the server as well
[20:06:12] <_newb> what is the best way to query if a document exists (return the id) before doing an insert?
[20:13:01] <_newb> what is the best way to query if a document exists (return the id) before doing an insert?
[20:53:21] <sshaginyan> I want to do something like this. db.test.update({ domain: 'atest.com'}, { $push: { technologies: { name: 'jquery', $push: { verifications: { proof: 'thisthat', source: 'http://www.google.com/' } } } } }, { upsert: true })
[21:11:50] <asido> I need to query documents by taking an average of 2 attributes in the document and compare whether it's within the bounds I want. does this sound like a doable thing with one query?
[21:18:35] <insanidade> hi all. what's the best cursor function for finding out the current size of an 'ever growing' collection?
[21:19:01] <tscanausa> your id in group would be the document id and then you would your "total" field as the $avg of the 2 fields
[21:58:02] <asido> is it possible to do something like this: db.coll.aggregate({ $match : { AttrInt1+AttrInt2: { $gt : 0, $lt : 100 } } }) ?
[23:33:51] <codenado> hi all. I have a question about many-to-many relationships. I have users and groups, users can be in many groups, groups contain many users. What’s the preferred way to store that? One option is a person list within the group document, and a group list on the user document, but that means data is duplicated