Log file Viewer
#mongodb logs for Thursday the 27th of June, 2013
[00:45:59] <bobinator60_> is there a way to $order_by a result set when the sort key is in an array like this: https://gist.github.com/rbpasker/68cd72645f7a51c7d0e9 ?
[02:19:07] <stephenmac7> Assuming the data in a document is not sensitive, is it secure to output the entirety of a document in an API (including ids and id references to other, possibly sensitive documents)
[03:18:23] <timetocode> i've got a little matchmaking server with a table of players waiting for an opponent, and when a new player tries to find a game they take an opponent from the list of waiting players (i then delete this player from the list of waiting players) -- but because reading is nonblocking, and writing is blocking, does this imply that the 'next available player' could be read multiple times before I successfully delete it from the tab
[03:20:15] <timetocode> i was thinking that if i performed an arbitrary change onto the player before deleting it, this would lock that particular record, but im not sure if thats the mongo-ish way to get things done
[07:59:10] <foofoobar> so I have a "jobs" table. The most queried by fields are: location (2dsphere), author._id, category. I plan to create an index for them by using: db.jobs.ensureIndex({location:"2sphere", category:1}
[08:52:11] <Sverre> is there any way to remove a document using mongoimport? Like, if I import a.csv, remove one line and then import it again the document will be deleted?
[08:54:28] <foofoobar> so I have a "jobs" table. The most queried by fields are: location (2dsphere), author._id, category. I plan to create an index for them by using: db.jobs.ensureIndex({location:"2sphere", category:1}
[09:03:34] <aladdinwang> django integrated with mangodb with basic python driver or use django mangodb engine, which is better?
[11:49:00] <tairov> Do anybody know how much time can it take to compact ~400GB database? If actual used space is about 100MB.
[12:52:29] <scanf> on an unknown deployment, whats the best way to get replication info and general info?
[13:16:19] <sdfghjhgffd> with this command: b.songs.findOne({origin: 'http://www.myzuka.ru'}, sort={aux_id: 1})
[13:16:54] <sdfghjhgffd> but its apparent equivalent in a python script: db.songs.find_one({'origin': origin}, sort=[('aux_id', -1)])
[14:27:27] <harenson> I mean, in the first query you get the document, then analyze it with Python or whatever you are using, then make the second query
[14:27:29] <revoohc> In a 3 member replica set, is there a way (without doing multiple stepDowns) to fail over to a specific host?
[14:29:43] <harenson> mift: maybe later you learn to do it in a better way, but, for now, it could help you
[16:54:05] <mrapple> so, i have a collection that contains user stats, calculated via map/reduce and dumped into a collection
[16:54:18] <mrapple> i want to allow sorting by different time period/public families/stats, however, mongo apparently has a cap of 64 indexes per collection.
[16:56:58] <millun> with mongotemplate on executeFindMultiInternal() i get IllegalArgumentException: can't serialize class
[16:57:01] <starfly> mrapple: typically, it's best to segment the data in that case, if that's possible
[16:58:10] <starfly> mrapple: you can move resulting collection into multiple collections; yes, segmenting has it's benefits and costs
[17:02:52] <mrapple> starfly: i've just read this for a bit, but i don't think aggregation would work
[17:03:48] <starfly> mrapple: it has limitations, but the only way to answer the question is to probe those limits with your data
[17:07:59] <starfly> mrapple: yes, on map-reduce, try aggregation to see if that can substitute (best I can offer)
[17:55:46] <n06> does anyone know of a chef cookbook out there that I can use to bake a mongodb clusteR?
[18:07:31] <millun> morning. i got a query in Java. specifically: Query query = new Query(Criteria.where("active").is(criteria.getActive()).andOperator(Criteria.where("animal").in(animals) where animals is @DBRef
[18:08:02] <millun> however, i can't seem to be able to execute 'find' with that query because Animal class can't be serialized.
[18:09:09] <millun> my guesses are running out of ideas. i am using spring-data-mongo. I don't have @Indexed on that @DBref and that @DBRef contains another @DBRef
[18:18:17] <Kzim> i have a problem, i miss an admin user on my mongos everything is good on each mongod but can't log on the mongos. is there a way to fix this without stop the service please ?
[18:53:10] <duncancmt> Hi! How can I print "ugly" json from the mongo shell? I've seen this http://stackoverflow.com/questions/14752450/mongodb-print-json-without-whitespace-i-e-unpretty-json but I really don't want to try removing whitespace with a regex
[19:02:13] <SpNg> I'm working on setting up MMS on an AWS VPC. My database server is not accessible outside of my VPC. What is the best way to setup MMS from one of the other nodes on the network?
[19:05:42] <SpNg> tystr: I have the database server setup to only communicate on my VPC, I have a fileserver and a web server that can communicate with the outside world. if I'm not mistaken, MMS needs access to the mongo log files. Is there a way to use MMS by connecting to the mongo daemon in my network?
[19:06:43] <starfly> Wish 10gen would allow local installations of the MMS stack, as they originally indicated they would
[19:06:43] <tystr> I *think* you can run the agent anywhere, as long as it can talk to your mongo instances and 10gen's url
[19:08:47] <SpNg> tystr: that would be great. I have not found the documentation for how to set that up. I have only seen the line nohup python agent.py > /[LOG-DIRECTORY]/agent.log 2>&1 &
[19:13:55] <SpNg> tystr: I feel like I would need to point the agent to the location on my network to connect to mongo. I will give this a run real quick and see what happens
[19:14:49] <tystr> "You can run the agent on any system that can connect to the MongoDB instances you want to monitor. As long as it can connect to each instance, you can use a single agent to do all the monitoring. Do be sure that the agent can make outgoing connections via HTTPS on port 443."
[20:15:30] <SethKniceley> hi, everyone! I'm a Mongo n00b, and was wondering if someone could help me out with a question on date storage.
[20:16:00] <SethKniceley> I'm working in PHP, and the app I'm working on will require searching by dates and times for events that happen in the system
[20:17:14] <SethKniceley> I was planning on storing them using Date, but someone was telling me that that was a mistake and that I should only store dates in seconds since epoch.
[20:23:11] <likarish> I like to use Date objects because they're human readable. Makes debugging easier but there may be reasons to use seconds since epoch too.
[20:24:44] <SethKniceley> I'm thinking I just create an index on the date fields I want to search by
[20:26:03] <SethKniceley> Just wondering why my friend was so opposed to me storing as Date and not seconds since epoch
[21:11:15] <Skunkwaffle> If anyone has a second, I'm trying to figure out if it's possible to convert the output of an aggregation pipeline, which currently looks like: [{'value': 1, 'key': 'one'}, {'value':2, 'key': 'two'}, ...], into an object so it looks like: {one: 1, two: 2, ...}. Can anyone help explain to me if/how this can be done?
[22:26:43] <tystr> primary is chugging along at ~30-40% cpu usage, and we're expecting an influx of traffic in a few days
[22:27:46] <tystr> our application is quite write heavy, so I suspect that's the cause, but I'd like to know for sure