Log file Viewer
#mongodb logs for Friday the 12th of June, 2015
[02:54:23] <sabrehagen> if no sort option is provided, does mongo db return results sorted by insertion date?
[02:59:35] <joannac> sabrehagen: if you don't provide a sort order, there are no guarantees about what order you get things back
[03:06:11] <rajat_> I need help with an aggregation , query : http://pastebin.com/sFuDZjpJ. I am getting this error "SyntaxError: Unexpected token : " Please help P.S. i am new with mongodb
[03:07:33] <rajat_> i have tried to run query individually, nd it'w working but here it is showing me this error .
[03:13:32] <joannac> rajat_: you don't actually understand JSON or javascript. there's not a lot I can do for you here
[03:15:32] <rajat_> ohk let me explain i am trying to merge http://pastebin.com/jzUN59tt and http://pastebin.com/TVWd0fNg both .
[03:18:19] <Boomtime> the trouble rajat_ is the two later examples you've provided contain valid JSON, where-as the example you provided at first contains, quite basic, syntax errors.. which i pointed out
[03:19:49] <Boomtime> i feel that you are trying to fly when you haven't learnt to crawl - you should start with simpler use cases
[03:21:40] <rajat_> I have only started learning it i will try to implement it from basic again . BTW thanks a lot
[03:22:31] <Boomtime> make sure you understand the rules of JSON before trying to construct complicated aggregation pipelines
[03:23:06] <Boomtime> the issues you have are not related to mongodb (at the moment) but with the syntax of the language you are using (javascript)
[09:51:02] <lqez> kas84: http://docs.mongodb.org/manual/faq/fundamentals/#how-do-i-configure-the-cache-size-for-mmapv1
[09:51:20] <lqez> If you're using WT, then set http://docs.mongodb.org/manual/reference/configuration-options/#storage.wiredTiger.engineConfig.cacheSizeGB
[10:01:27] <kas84> It’s definitely loading that conf—> ps -ef |grep mongod —> mongod 1889 1 26 09:42 ? 00:04:18 /usr/bin/mongod -f /etc/mongod.conf
[10:11:15] <lqez> I ran lots of aggregation and other jobs on same instance so it uses more than 32G that I set
[11:20:54] <Hanumaan> getting following error from the server "unauthorized db:tapari lock type:-1 client:156.211.156.175 any solution?
[14:43:18] <jr3> https://gist.github.com/anonymous/3436b47da1592169af9e -- If I comment out this populate I go from 1.5second return to a 300ms return
[15:30:03] <csd_> Is there any way to create something similar to a tailable cursor for a non-capped collection?
[15:35:26] <csd_> saml: I have a query that I basically want to create a listener for. It's basically like, give me the result where attr is smallest. And so listener would then let me know immediately when this changes
[15:38:22] <saml> actually, doesn't have to be heap. in your script, you'll store an _id and attr. and it tails oplog. for updates or removes, compare attr.. and update _id and attr accordingly
[15:40:32] <saml> http://stackoverflow.com/questions/20871009/how-can-python-observe-changes-to-mongodbs-oplog
[15:41:01] <csd_> the find() is working over a set of tasks that have a time attribute for when they need to be executed and another attribute saying if the task is in process / completed. so it might be that the task with the nearest time isn't ready for execution yet, so youd have to wait . but during that waiting period, a new task with a smaller time could be added, and that would have to be executed first
[17:47:34] <rajat_> can these both queries can be merged in the following way http://pastebin.com/sMWQSbhv and http://pastebin.com/7Hn1A5FP and my solution is http://pastebin.com/tUzVEreL please help
[18:13:23] <symbol> haha sure - I'm just tired of seeing the first result on google being that "Why you should never us MongoDB" post
[18:14:41] <StephenLynx> if one just see a random blog post and decides without concrete arguments, its their loss.
[18:15:20] <GothAlice> https://blog.serverdensity.com/does-everyone-hate-mongodb/ < a good article about other articles, and the SD guys have many other great articles, too.
[18:15:55] <symbol> I agree, that FOSS mentality is dangerous and I suppose it makes sense that the controversial posts get the higher rankings.
[18:16:13] <GothAlice> They're more heavily linked to, esp. by those wanting to counter the arguments.
[18:16:32] <GothAlice> Google "teaches the controversy" rather than providing true facts, on most searches. ;)
[18:19:09] <GothAlice> The vast majority of my FOSS code is actually commercial FOSS. I was paid to work on it.
[18:20:03] <GothAlice> FOSS is a value-add for business due to the ability for improvements to be contributed at little cost (code review time, basically,) and community recognition.
[18:21:24] <symbol> I love hearing about companies that pay devs to work on FOSS (now that I know the correct definition)
[18:22:47] <GothAlice> Our philosophy at work: if it's not business critical, open it up. The secret is not about the bits, it's about how the bits are specifically put together.
[18:27:10] <GothAlice> For example: a distributed MongoDB-based RPC system is nifty and all, but open-sourcing it reveals little about the things we're using it for, such as screen scraping or data processing (fully proprietary code).
[18:52:27] <m3t4lukas> is it just robomongo or is something like "db.getCollection('test').find({"array": {"$size" : {"$lt": 8}}})" not possible right now?
[18:59:37] <GothAlice> m3t4lukas: You might need an --authenticationDatabase argument, if your user is in a different database than the one they have access to.
[19:03:30] <m3t4lukas> on my test server there is a database test with a user "lukas" with the password "lukas"
[19:05:46] <m3t4lukas> okay I tried "db.getCollection('test').find({"array": {"$size" : {"$lt": 8}}})" now
[19:20:28] <m3t4lukas> if you want to test your index just do some small code in any language you prefer that a driver exists for and generate a massive amount of documents and test your queries with and without an index and compare those results
[19:28:26] <brotatochip> hey guys, I want to create a database user in mongodb for Icinga to be able to use to monitor the replication status, can anybody tell me what roles I'll need to grant that user?
[19:29:55] <brotatochip> basically I want the user to be able to execute rs.status() on the command line(similar to mysql -e) so I can grep/awk out the relevant info
[19:34:52] <brotatochip> oh actually I think I found it, but feel free to correct me if I'm wrong: db.createUser({user:"icinga",pwd:"<password>",roles:["replSetGetStatus"]})
[19:45:59] <brotatochip> StephenLynx: Icinga is just a fork of nagios and entirely irrelevant to what I'm trying to do :-P
[19:46:43] <brotatochip> Ah, it's a monitoring server, executes checks and sends out emails if a condition is met or not met
[19:48:04] <brotatochip> I'm trying to create a custom role to grant access to the replSetGetStatus action on the cluster resource
[19:54:08] <brotatochip> StephenLynx: Kibana is the gui for the ELK bundle - Elasticsearch, Logstash, and Kibana
[19:57:17] <brotatochip> The documentation for this page seems to be broken http://docs.mongodb.org/v2.6/tutorial/define-roles/
[19:57:32] <preaction> Square was building something like kibana for mongo, i don't remember what it was called though
[19:58:18] <brotatochip> Specifically this command: db.createRole( { role: “replicaMon", privileges: [ { resource: { cluster: true }, actions: [ “replSetGetStatus" ] } ], roles: [] } )
[19:59:05] <brotatochip> Would anyone be able to tell me how to properly formulate that query because the one in the doc does not work
[20:20:27] <brotatochip> anybody around who can help me understand how to configure a replica set with auth enabled?
[20:32:19] <brotatochip> btw I think my syntax errors when attempting to create the custom role was due to weird characters being copied into my terminal from the notes app that I was using for editing the command...
[20:55:26] <daidoji> Hello, when I create an index does it immediately show up in getIndexes or does it wait until its completed its creation?
[21:12:32] <brotatochip> cheeser do you know if I modify my production standalone mongodb instance to a replica set, add my secondaries, and the initial sync begins, will I be able to bring my application back online or should I schedule a downtime for however long that sync takes?
[21:14:43] <brotatochip> or deathanchor do you know the answer to the question I just asked by chance?