[00:09:24] <tjmehta> Hi all, quick question: If create a compound index like { foo:1, bar:1 }. Will that also create an index for {foo:1}, so I don't have to index that field outright?
[00:11:40] <joannac> no, but searching on foo will be able to make use of that index
[00:11:50] <joannac> so you don't need a separate {foo:1} index
[00:12:33] <cheeser> but you will need one to query against just bar
[00:53:58] <tjmehta> joanac, another indexing question for you: http://www.codeshare.io/NMKzU
[00:54:50] <tjmehta> Are indexes used for $elemMatch queries? In the link I send you you can see I am find a project by id and then using an $elemMatch to find a particular file.
[01:34:10] <epx998> Is there a decent howto guide for upgrading from 2.2 to 2.6?
[01:49:35] <engirth> suppose you have a document that has a changes [] that you push into changes for a given day. I'd like to get the array index for $push since I want to maintain a reverse lookup.
[01:54:53] <dpg2> will mongo ever delete a record on update?
[01:57:35] <dpg2> nevermind. it just... replaced everything.
[03:47:00] <blizzow> I want to create a query for all documents in a collection inserted after ObjectId("l38463358349slj2398r"). Anyone can someone here explain how I'd do that?
[04:12:25] <lacrymology> is there a way to not insert duplicate objects, when not passing an _id? I mean, besides doing a find for each object I want to insert
[04:15:49] <x0f> lacrymology, do you need to update this document? otherwise add an index with a unique-key, which raises an error upon insertion
[04:16:18] <lacrymology> x0f: what if I'm inserting a list of objects and one fails?
[04:17:07] <lacrymology> I mean, will the whole insert fail, just the one object, or from that object on in the sequence?
[04:18:39] <x0f> depends on your method of insertion, i guess. afaik transactions are atomic so the single insert for this document will fail, how your driver handles it depends if you need to rerun ist, or whatevers.
[04:26:11] <kataracha> is there a recommended production setup for connecting to mongodb server on a local network? ie should I stay listening on local interfaces and create an ssh tunnel or listen to external interfaces directly and give my connecter the private ip etc.
[15:38:33] <chiku> Node Mongo dB inserts fail siently
[16:41:12] <kataracha> has anyone had experience building and mainting mongo themselves here? was there significantly more overhead compared to using a package?
[16:49:02] <kataracha> what about patches etc or is it mostly self sufficient for a long enough period
[16:50:54] <kali> we were re-building at each new release
[16:50:54] <kataracha> mostly I was interested in using --ssl is all and either I would have to build mongo or use something external like a ssh tunnel or stunnel and am wondering the best route re overhead/performance
[16:51:31] <kali> tunnels will probably not work with anything more complex than a standalone server
[17:19:46] <kataracha> kali: what about setting up the mongod service to run with the recommended ulimit values and upstart scripts etc
[17:20:06] <kataracha> was there much peripheral setup after the build
[17:21:55] <kataracha> or if I already have the 2.6 package install can I mostly just build mongo how i wish and drop in the binary
[17:22:50] <kali> kataracha: just make sure you set the package on hold
[17:22:59] <kali> kataracha: another option is to rebuild the package
[17:29:02] <kataracha> I dont understand why ssl isnt just build into the packages to begin with is there a reason other than adding value to the enterprise service
[17:29:52] <kali> kataracha: i don't know for sure, but it might be the paranoid american exportation laws
[17:30:17] <kali> (and... I've just been flagged by the NSA)
[22:01:21] <felipesabino> can anybody help how to setup some databases with replica set enabled using docker (and may be fig)? http://stackoverflow.com/questions/27449306/what-is-the-proper-way-of-setting-a-mongodb-replica-set-using-docker-and-fig