Log file Viewer
#mongodb logs for Tuesday the 13th of October, 2015
[01:20:28] <fructose_> I'm looking to have a field that keeps a count of items, ideally so I can write a query with an array of those items and the count will increment. Can I do that in a single update?
[01:23:43] <markizano> When mms does backups, is there any component run externally from the server running the mms agent (e.g. from mms.mongodb.org?)
[01:24:02] <fructose_> e.g. if I run an update with {id: myId} and (somehow) items: ["apple", "banana"], it will store "apple": 1, "banana": 1. Then if I did it again with just ["apple", "carrot"], it'd end up storing "apple": 2, "banana": 1, "carrot": 1
[01:24:09] <markizano> we are trying to do a rs backup, and it's unable to connect, curious if the target's DNS is resolved ...
[01:30:16] <fructose_> joannac: I need it to be a field of that document. So would it be update({}, {$inc: {items.apple:1, items.banana:1}) ?
[01:33:44] <joannac> fructose_: then your example is wrong. what do you actually want the target document to look like?
[01:34:53] <joannac> fructose_: you said "it'd end up storing "apple": 2, "banana": 1, "carrot": 1" but what do you actually want? {items: {"apple": 2, "banana": 1, "carrot": 1}} ?
[01:35:43] <joannac> sorry, to make it more clear, {_id:1, items: {"apple": 2, "banana": 1, "carrot": 1}} ?
[01:36:01] <claytonzaugg> Any MeteorJS users out there? If I have an order form and I need the order number to automatically increment each new order created, how can I do that?
[01:52:27] <claytonzaugg> Any MeteorJS users out there? If I have an order form and I need the order number to automatically increment each new order created, how can I do that?
[02:28:24] <iio7> I have just done some testing with mongodb, disabled the daemon and then enabled it again, but now it wont start: http://pastebin.com/ux4wdjCA
[02:31:27] <iio7> I just did with "-vvvv". It started. Doing some JOURNAL durability "groupCommit begin", processing commit number 1025..
[03:28:57] <rjurney> I need to look at records in the links collection that have values other than ‘partnership’ and ‘customer’. How do I do that? This isn’t working: db.links.find( { $or: [ { type: {$ne: 'partnership' } }, { type: {$ne: 'customer'} } ] } )
[03:31:46] <rjurney> I figured it out: db.links.find( { type: {$not: {$in: ['partnership','customer'] } } })
[04:03:17] <Boomtime> for your case it's probably the same, though be aware that if $type is an array, $nin and $not:$in will produce different outcomes
[04:09:24] <Bioblazin> First time doing mongodb, and I would like to hear a few pointed on a Schema Design. http://pastebin.com/HneDYDWK
[04:10:07] <Bioblazin> the Schema Reflects a Product API Key Storage, associated with a Person who created it.
[04:10:40] <Bioblazin> :x Is it better to generate a Associtive GUID? Or use the _id of the Entry for Searching?
[09:54:15] <Creeves> hey guys, I'm getting error code 17406 (getMore executor error due to trying to sort in ram and running out of space) and read online that puttin an index on the field sorted by should fix it
[11:09:25] <Creeves> cheeser: it was a .find({_id: {$in: some_array}).sort({some_field: 1}) query, and i put an index on some_field but it still gave that error
[11:10:28] <Creeves> now it's something like find({some_field: {$gte: start_index, $lt: start_index + limit}}).sort({some_field: 1})
[11:47:37] <amitprakash> Hi, I am using mongodb to insert records from collection A to collection B as db.A.find(filters).sort({_id: -1}).forEach(function(row) { db.B.insert(row); });
[11:48:00] <amitprakash> However, db.B.find().count() returns a single record ( when the insert has been going on for a while.. what gives?
[14:34:38] <amitprakash> Why does pymongo keep throwing a pymongo.errors.ServerSelectionTimeoutError: No replica set members match selector "<function any_server_selector at 0x7f5a6c05f378>"
[17:05:09] <parshap> How do you get the _id of a newly inserted document when using the node.js mongodb-core driver? The documentation doesn't mention anything and the result object doesn't seem to have an _id in it. http://mongodb.github.io/node-mongodb-native/core/api/Server.html#insert
[17:17:05] <parshap> StephenLynx: seems to be the suggested driver to use. mongodb-native is referred to as "legacy"
[17:17:30] <parshap> StephenLynx: mongodb - MongoDB legacy driver emulation layer on top of mongodb-core
[17:18:01] <StephenLynx> The MongoDB driver is the high level part of the 2.0 or higher MongoDB driver and is meant for end users.
[17:18:35] <StephenLynx> https://www.npmjs.com/package/mongodb-core: Core MongoDB driver functionality, no bells and whistles and meant for integration not end applications
[17:20:40] <parshap> StephenLynx: i wanted to use the lower level api intentionally, but i'll give mongodb-native a try
[19:37:27] <Bioblaze> I recently just began working with Data Modeling, I come from a Mysql/Spreedsheet style background.
[19:38:36] <Bioblaze> When it comes with Data Modeling, its obviously diffrent from Standardized Data Association much like Table Association in Mysql.
[19:38:58] <Bioblaze> Whats the most Optimal Way to Handle Data? Should it be stored in the Data Structure of the Data Model?
[19:39:52] <StephenLynx> I heard from an experienced dev that using multiple databases at the same time doesn't perform well. as in, multiple databases connected running under the same server.
[19:41:55] <StephenLynx> the real question one should make is to make references from a collection to another or nest the data.
[19:42:33] <StephenLynx> and theres no absolute right answer to that, is depends on how you are going to query and what load you expect to handle.
[19:43:20] <StephenLynx> if the nested object would be too complex and I expect an unlimited amount of them, I go for relations.
[19:44:06] <Bioblaze> Its a Product Handler for Virtual World Products. Storing the Associated Creator, and then Security Tokens for Inquires too the Server.
[19:47:46] <StephenLynx> IMO, theres no point in adding schema validation if the database dont actually enforce it.
[19:49:26] <StephenLynx> which increases RAM consumption, since every object now packs much more than just it's data and is not how application code is meant to be used in js.
[19:52:01] <Bioblaze> I have Associtive Avatars, Products, Regions, Parcels, and a Collection for Traffic more or less.
[19:52:32] <Bioblaze> The productSchema I showed you in that link, is ment to store security skys, and associate the product with the creator.
[19:53:18] <StephenLynx> if you put it into a separate collection, now you have to perform two queries to retrieve
[19:53:41] <StephenLynx> while if you embed, you don't have to perform any aditional query for this extra data.
[19:54:13] <StephenLynx> if you have to perform complex updates and projections, embedding will put some limitations on you and the code will be more complex regardless.
[19:55:37] <Bioblaze> Can i explain too you my System, and you can give me some advice on structuring/embedding?
[19:55:40] <StephenLynx> and takes some time to learn about mongo's strengths and weaknesses to be able to decide on that more correctly.
[20:02:23] <Bioblaze> XD thx u Derick I just didnt wanna typoe abuncha stuff and be told "Breaking rules" XD and kicked.
[20:04:29] <Bioblaze> So far I have about 3-5 Collections, Product, Avatar, Region, Parcel, and Traffic.
[20:04:43] <Bioblaze> Product contains information about things created by Avatar, for Security keys.
[20:05:11] <Bioblaze> Parcel is Associated to Region, and then Region is Associated in Traffic, along with Avatar associating the Avatars who have Visited.
[20:05:47] <Bioblaze> Avatar, Region, Parcel are all Found at Randomly, and then Sent to the Associated Collection.
[20:06:30] <Bioblaze> Looking at about hmm. 500-1000 calls to the Database in total per half day between all the systems.
[20:07:03] <Bioblaze> What i`m having issues with. Is my Associative thinking too Mysql, and using that in Mongodb which is obviously wrong in its attempt.
[20:07:32] <Bioblaze> So i`m looking to learn Proper Structure of a Collection, or the theory of a Proper Developed Collection structure so I can work on mine oh. Lulz.
[20:11:00] <StephenLynx> another detail about nested documents: if you ever find a situation where you don't want to restrict yourself for a single parent object, it will be harder to query and more expensive.
[20:11:14] <Bioblaze> But yea. I`m trien to figure out the proper development of these collections as of right now XD
[20:12:53] <StephenLynx> doing that would be much harder if threads were stored inside the board they live.
[20:14:56] <Bioblaze> ProductSchema should prolly be inside of Avatar, and Parcel should be inside of Region, then I simply Associate Region to Traffic. o.o
[20:15:36] <Bioblaze> only queried entries for viewing are Traffic. and Checking of the Product Key.
[20:15:40] <Derick> Bioblaze: you need to think not how to structure your data, but how your application uses and interacts with data - then according to *that*, you will find out what your schema should be
[20:17:33] <Bioblaze> Well Currently, there is only 1 Check preformed for the Security, then a Token is exchanged, then everything else from there is pushing the Current Location to the Server for Traffic nothing else actually.
[20:24:16] <Dizzle> This is probably a dumb question, but in the latest version of Motor, do you even need to install tornado if you intend to use asyncio?
[20:30:13] <StephenLynx> you are the first guy mentioning motor or tornado in this channel I have ever seen.
[20:31:39] <Bioblaze> StephenLynx if I do a Parcels: { Name: String, Desc: String } it will create a Id inside of the Collection and inside of the Parcels Array Correct?
[20:32:20] <Bioblaze> so both Region, and the Array inside of Region.Parcels will both receive a ID?
[20:35:46] <Bioblaze> http://pastebin.com/jWbxy6bm this is my current data outline :X remake of Regions.
[20:36:21] <Bioblaze> Only issue I am attempting to figure out, how do Associate the Parcels with another Collection.
[20:36:53] <lrvick> Can anyone think of a way to pipe a password to a mongo cli string in such a way that it does not show up in ps, etc?
[20:36:53] <Bioblaze> Xd what do you mean? Description of each Piece or the total amount of the collection explaining what its for.
[20:42:35] <edrocks> can you use multiple position operators to delete fields from multiple arrays?(not nested just several at the top level)
[20:44:44] <StephenLynx> so you have an array named "Name" which contains objects that have a field called "Name"?
[20:46:22] <Bioblaze> I`m just not sure if I should put more Data into it, since "Names, Descriptions" Changed constantly.
[21:28:50] <Derick> Hey! We're running a driver survey at MongoDB. If you have a few minutes, please fill in https://www.surveymonkey.com/r/52Z9SXF
[22:23:48] <rangergord> Hi. I'm looking at running Mongo on an ARM embedded device (relatively beefy, 800MHz and 512MB RAM). The page http://docs.mongodb.org/manual/administration/production-notes/ seems to strongly indicate you need ext4 or XFS as a filesystem. What issues can I expect to have if I try to run it on a jffs2 filesystem?
[22:24:56] <rangergord> the page also recommends Linux 2.6.36 or later. My system is at 2.6.35, what unpleasantness can I look forward to?
[22:34:32] <happyken> rangergord It just means they haven't tested jffs2 file system. They did tested ext3 and ext3 doesn't have fsallocate system call. It has to do padding to allocate space which might take couple of seconds. So they don't recommend ext3. So jffs2 file system might work fine for you.
[22:36:03] <happyken> rangergord It should work for you even with old kernel but might have performance impact. You can try and post your results for the community.
[22:36:21] <happyken> I have mongo db with multiple databases having common prefix. I want to have common account credentials for dbs having common prefix. Is this use case possible by any chance ?
[22:40:06] <inorganik> hi there, was hoping someone might be able to give me some mongodb schema advise
[22:49:37] <steffan> inorganik - if you provide a brief description of your problem it may encourage feedback
[22:54:46] <inorganik> I want to store a collection of podcast episodes, but within that user interactions with episodes... so I envision an episodes collection with array of users
[22:55:48] <steffan> It's not necessarily a bad design, but it would depend on how large you expect the arrays to get
[22:58:19] <steffan> Okay. So you're using the MMapV1 storage engine. MongoDB will have to move documents once they grow past their allocation, which will affect performance slightly.
[22:59:23] <inorganik> so, for instance a new comment by a new user... is it possible to write an upsert
[23:00:55] <inorganik> but i don't want to add a dupe user - all interactions by that user should be under that user
[23:03:09] <steffan> Could you do that as two separate operations? The creation of a new user should happen much less frequently than creation of an interaction
[23:06:30] <steffan> Do the users have a relationship with the episodes themselves, or only if they create comments / interactions?
[23:07:03] <steffan> If that is the case, then it probably supports moving user metadata to its own collection and storing an array of comments+user display names with the episode
[23:11:22] <inorganik> so maybe: db.episodes.update({title: 'some episode'},{$addToSet: {users: { username:inorganik }}});
[23:11:29] <inorganik> then db.episodes.update({title: 'some episode', users:{ $elemMatch: { username:inorganik }},{ 'users.$': { $addToSet: { events: { type: 'comment', comment: 'some comment'}}}})
[23:26:15] <meye1677> I have been getting gpg errors when installing mongodb-org on ubuntu trusty running the commands in the docs http://docs.mongodb.org/manual/tutorial/install-mongodb-on-ubuntu/
[23:27:51] <meye1677> I’ve had those commands in a script and successfully install every day with the same script—I started getting the gpg errors today.
[23:44:33] <Bioblaze> Should I store Ownership of a Product within the Products Schema?? Or within the Account Schema?
[23:53:59] <steffan> Bioblaze: Consider the relative quantities. What are the numbers of products & accounts?
[23:54:50] <Bioblaze> umm Products could be maybe few 100 per Account, but could be 1000s of Owners per Product.
[23:56:25] <Bioblaze> right now the Accounts Table consists of Name, broken down virtually into First/Last and a UUID along with a creation date. So almost no information.
[23:57:31] <Bioblaze> Prolly should just store the information in the Accounts for the Products they own. Since the 16mb limit