Log file Viewer

[00:44:37] <tystr> hmm

[00:44:58] <tystr> with map/reduce, emit isn't "grouping" like it's supposed to

[02:27:05] <nemosupremo> okay I have a problem. I have 1 machine, in a replica set thats secondary. The other machine went down, but unfortunately was given a bad hostname so it can no longer connect to the first machine. I naively took the arbiter down before I realized I had this problem because I wanted to change its configuration. How can I reissue commands to rsconf on a secondary machine?

[02:32:20] <nemosupremo> figured it out {force : true}

[07:22:10] <EricL> Anyone have any experience upgrading from one to three config servers in a sharded environment?

[07:22:33] <EricL> Keep getting this error and Google isn't much help: "$err" : "could not initialize sharding on connection ip-10-5-25-231.ec2.internal:27016 :: caused by :: specified a different configdb!",

[07:28:02] <[AD]Turbo> hola

[07:34:17] <balboah> EricL: I got that once, but I have no idea why it was. Maybe different hostnames in config and on server or something

[07:35:09] <balboah> EricL: but if the other servers are alive, shouldn't it get re-populated if you would wipe it out

[07:37:00] <EricL> We've tried that.

[07:37:54] <EricL> We've tried killing everything and re-starting.

[07:38:02] <EricL> Still a no go.

[07:38:03] <balboah> I got it once when trying to move a config server, I just re-copied the init script and it got fixed

[07:38:33] <EricL> We have like 40 mongos servers.

[07:45:51] <balboah> EricL: make sure --configdb specifies correct hostname as seen by the config server. I get that for example if I change the hostname to localhost instead of the specified hostname on the server

[07:46:25] <balboah> (both mongos and mongoconfig run on the same machine)

[08:07:00] <hazard_> does mongo support $in queries on shard keys and does if so does it do so by querying only the necessary shards in parallel

[08:31:29] <Bartzy> Hey

[08:31:38] <Bartzy> Anyone running a big MongoDB setup on VMs ?

[08:31:41] <Bartzy> For example, Linode or AWS ?

[08:33:10] <c0815> lots of people do

[09:31:20] <MaxGaudron> Hello

[11:27:52] <PDani> hi

[11:30:40] <PDani> i have a collection which have documents like {a:3, b:4, payload: "foo"} where a and b are integers. i'd like to query with a single bson (one network roundtrip) every distinct a's with the greatest b value. is it possible?

[11:32:17] <NodeX> with a map/reduce

[11:33:05] <NodeX> you can use a query with distinct but you cant do MAX(b) - like you may do in SQL

[11:36:26] <PDani> i suppose sharded collection excludes group() as well...

[11:36:34] <kali> PDani: they do

[11:37:00] <PDani> ok, thanks

[11:48:01] <remonvv> PDani, if you're 2.1+ you can do it with a single roundtrip through the aggregation framework. All other options available to you are suboptimal (not a single roundtrip, JS context based and thus slow, etc.)

[14:00:25] <PDani> could somebody give me an example of the usage of $first operator in aggregation framework?

[14:01:52] <FerchoDB> if I have a collection of documents which have an array of integers. Is it possible to search documents in which any of the numbers in the array is greater than X?

[14:07:28] <NodeX> {"foo.number" : { $gt : X}}

[14:13:14] <PDani> db.testc.aggregate({$sort: {b:1}, $group: {_id: "$a", b: {$last: "$b"}}}), why I get the same results as in db.testc.aggregate({$sort: {b:-1}, $group: {_id: "$a", b: {$last: "$b"}}})? I reversed the sort order, so the last seen element couldn't be the same...

[14:13:47] <PDani> db.testc.find(); { "_id" : ObjectId("4ff44840f1de0f7cb11aa88a"), "a" : 1, "b" : 1 }; { "_id" : ObjectId("4ff44843f1de0f7cb11aa88b"), "a" : 2, "b" : 1 }; { "_id" : ObjectId("4ff44847f1de0f7cb11aa88c"), "a" : 1, "b" : 3 }; { "_id" : ObjectId("4ff4484af1de0f7cb11aa88d"), "a" : 2, "b" : 2 }

[14:13:51] <PDani> this is my collection

[14:14:39] <FerchoDB> thanks NodeX, I was trying that but was missing a {

[14:14:48] <FerchoDB> now it worked :)

[14:20:12] <hpg301> Hi... I am looking for a nosql db to replace one now implemented in bdbxml. I will be storing documents having a tree structure, like an xml doc. I'm new to MongoDB and have read some intro material but did not see how I can query a MongoDB db to get anything like an XPath query. I look forward to reading comments on how MongoDB could be used for my application.

[14:24:30] <c0815> MongoDB is not an xml database...if you want to store trees then look into a graph db or something similar. MongoDB is not very suitable here

[14:25:28] <balboah> hpg301: there are probably many things you can do in XPath that is not possible. An object can contain other objects in a tree-like structure, but it's not easy to query different "paths"

[14:26:18] <neil__g> @hpg301 you should store your tree to fit your data structure to the query and not the other way around

[14:26:32] <neil__g> that wasn't very good english

[14:27:03] <NodeX> you should always store data to make the query the most efficient

[14:27:11] <NodeX> everything else can be done app side

[14:28:57] <c0815> data should be stored in a reasonable consistent and coherent way....

[14:29:16] <c0815> adjusting your data model to ulimate speed is unlikely desirable

[15:18:48] <tazle__> has anyone experienced a replication failure where a secondary newly elected primary would continue attempting to replicate from another node and a secondary that formed the new cluster together with the primary would not attempt to replicate from the primary?

[17:24:38] <coopsh> what happens if a mongod --repair is started without --fork and the processed is interruppted (eg. ctrl-c)?

[17:24:52] <coopsh> is there a potential data loss? corruption?

[17:27:10] <c0815> unlikely since the original data should remain untouched

[19:29:44] <multiHYP> hi

[22:08:33] <W0rmDrink> does mongodb align data on page boundries in its data file ?

[23:58:51] <MarekkPie> anyone got time to answer a question about the mongodb/flask tutorial?