Log file Viewer
#mongodb logs for Friday the 28th of August, 2015
[00:30:30] <justinsf> Hello, is there anyway to speed up creating indexes, after running mongoimport? All of our indexes have { background: true }. Our server has 16 cpus, but noticed only 1 cpu is being used to index. Is there anyway to make creating background indexes multi-threaded?
[00:33:37] <Doyle> justinsf, indexing during initial sync drives me nuts. Same behavior. Get a DB, index, get a db, index, seemingly single threaded
[00:34:08] <Doyle> I know there was supposed to be an upgrade to initial sync between 2.6.? and 2.6.7 to enable 32 threads of replication, but I don't see it.
[00:36:42] <justinsf> Is there a configuration option I need to set in config to increase concurrently of creating indexes?
[00:39:24] <justinsf> from the doc: " The background index operation uses an incremental approach that is slower than the normal “foreground” index builds. If the index is larger than the available RAM, then the incremental process can take much longer than the foreground build."
[00:46:06] <Doyle> I like to run tmux with the following: top, iotop, iostat -xmt 1, iftop, tail -f mongod.log, mongostat, plus a pane for mongoshell
[00:57:04] <justinsf> 2015-08-28T00:53:33.236+0000 --authenticationDatabase is required when authenticating against a non $external database
[00:58:11] <justinsf> *0 *0 *0 *0 1 2|0 0 48.1G 97.0G 13.2G 0 0|0 0|0 300b 11k 27 brandcast PRI 00:54:5
[00:59:26] <Doyle> the qr|qw ar|qw are important columns. If your servers' getting slogged those values will just climb
[02:10:13] <Lnxmad> hey, anyone have advice on a "user" collection. storing username, hashed password, but also having an embedded "profile".
[02:12:00] <Lnxmad> I am contemplating, incase I ever want to offload the user collection to "more" secure db
[02:47:29] <quasiben> I'm working with an older version of mongo: 2.5 or 2.6? in any case, i'm helping to develop a web application to store 3rd party auth information and was looking for guidance
[02:48:42] <quasiben> So UserA has login credentials to the site and hey store credentials for digital ocean or some other service. Are there recommendations on how to store those 3rd party creds in away which is secure and can only be read by UserA
[14:16:53] <mbwe> doc_tuna: say i have a document like {name: "mbwe", age: 12} and if i update that document with exactly the same content for that document, i don't want that update go trough
[14:19:31] <doc_tuna> it's a no-op if you send it with the same contents but you are still wasting the databases cpu and network by sending the update, it would be better not to send it
[15:52:50] <hernan82arg> I've been researching a bit and I couldn't find any document how to do what I want to do
[15:54:02] <hernan82arg> and I want to configure an endpoint (DNS entry) that points to the primary node of the mongo cluster
[15:56:39] <Derick> no, don't try to outsmart the replicasets and leave it to the drivers to find out where the master is
[15:58:34] <hernan82arg> okay, so the application connecting to the cluster should know which node is primary then/
[15:58:36] <hernan82arg> okay, so the application connecting to the cluster should know which node is primary then?
[19:40:26] <cheeser> each line is a valid json document but collectively the file won't parse as a single json doc.
[19:45:58] <terminal_echo> through, csv export, rsync, BULK insert of csv with SQL and updating the info with temporary tables
[19:46:11] <terminal_echo> someone tell me that's crazy because it seems extremely simple and elegant to me
[20:16:30] <terminal_echo> maybe, but the export csv is streaming, the rysnc in streaming, and the bulk insert is quite fast..
[21:11:32] <saml> what's downside of using string as _id ? Most of my queries will be db.docs.find({_id: <url> })
[21:12:06] <saml> or i can leave _id alone and create a new field, url, which has unique index. not sure what's better
[21:12:26] <terminal_echo> cheeser: thanks for the sqoop idea, this is definitely the way to go, i take it its easy to transfer from mongod -> MSSQL?
[21:23:31] <MacWinne_> if I want to insert an activity document into a collection, but I want to not insert the document if a specified set of fields already match in the collection already, what would be the best way? I want to do this atomically... I'm not looking to upsert since I don't want to update the existing record
[21:26:11] <MacWinne_> also I can't use indexes for it beacuse the sets of fields I'm looking for can vary
[21:38:50] <diegoaguilar> hello I have documents in a collection with a onDemandPoints attribute looking like this
[21:38:51] <diegoaguilar> Ahorita te pague normal y lunes o martes espero poderte confirmar cuando cambiamos a asimilados
[21:40:24] <diegoaguilar> I must say that onDemandPoints is an attribute for each object inside a players array attribute
[21:40:56] <diegoaguilar> so document looks like {players: [{id: "", onDemandPoints: {totalPoints: 15}}]}
[21:42:17] <diegoaguilar> how can I find "documents with a document with value "something" and onDemandPoints with value x inside players array?
[21:58:37] <Nikesh> Well, I'm just considering which Ubuntu to install on a new machine. I've seen that Mongo doesn't officially support 15.04 yet, so I wanted to see if others got it to work
[23:01:16] <roo00t> i don't know mongodb right now but i want to use it with my RESTful API. Could someone tell how the json file looks like so that i can test my code
[23:13:44] <Boomtime> you have bought an empty filing cabinet and you are asking what is the the content of the pages that are put in it - there is nothing in it, until you put it there
[23:35:40] <roo00t> https://www.mongodb.org/downloads#development still doesn't have a stable version for 15.04