Log file Viewer
#mongodb logs for Monday the 30th of July, 2012
[04:09:51] <SalocinTEN> hi, i have a problem with a 3 shard cluster. i couold add the first and third shards but the second shard keeps saying that it cannot connect.
[04:10:34] <SalocinTEN> i am also using the ports 27107 for all the servers of each shard replicaset.
[04:12:26] <SalocinTEN> "errmsg" : "couldn't connect to new shard mongos connectionpool: connect failed
[05:58:54] <thestereobus> http://stackoverflow.com/questions/11714675/mongo-ruby-driver-cannot-connect-to-replica-set
[06:01:14] <adiabatic> it's around midnight on a Sunday/Monday. I'd recommend waiting 10 hours or so.
[06:23:29] <omid8bimo> hello. i have this error when i post some photos on my db after i switched primary masters in replicaSet
[06:23:50] <omid8bimo> <h1>MongoGridFSException</h1> <p>Could not store file: E11000 duplicate key error index: bimodb.image.chunks.$files_id_1_n_1 dup key: { : ObjectId('501505c8be98f0ff12000079'), : 1 }</p>
[06:27:50] <thestereobus> I figured out my replica set problem... about 12 hours of debugging. Will answer my own SO question
[06:37:48] <thestereobus> so I realized that mongo couldn't see the other node by the name I was providing
[06:49:08] <Guest76115> hello all..I am a newbie, and I wonder when will the version 2.2.0-rc0 become the production level? since I found some shard operations are quite enhanced in the development version..
[08:49:47] <Dreamer3> looking at this new thing, it sends a SIGINT to stop, but for me that still leaves a lock file
[08:56:56] <Dreamer3> so i guess if i move my mongod.conf that maybe this control panel thing will really work
[10:28:06] <hjb_> howdy. I#ve been requested to deploy mongodb on a test machine, but got stucked on the following question now: should i go for v8 or spidermonkey?
[10:31:42] <Gizmo_x> I got error like this: http://pastie.org/4357977 unauthorized using zend framework 2 and doctrine-mongo-db
[10:36:04] <Lujeni> Hello - It's dangerous to launch two mongodump on different collection ? ( shard environment) Thx
[10:48:17] <hjb_> NodeX: great. I feel thought, that i'd not use what i like, but what's more stable ;-)
[10:54:46] <hjb_> NodeX: i've found a quite fresh comment in gentoo bugzilla that suggests that upstream mongodb developers think that v8 isn't stable enough yet
[11:05:26] <NodeX> perhaps you shouldv'e posted your question in a different manner like "Is spidermonkey or v8 more stable" ;)
[11:09:58] <cromeno> I want to use mongo for storing visitor statistics like clicks, views etc.. and then use mysql for client data.
[11:10:52] <cromeno> and the common reference is the "object id", that is the id that every object (banner for example) has
[11:11:57] <neil__g> we have product IDs and we've used that with some success, so the mongo object just has product: <id>
[11:13:13] <cromeno> For example "Give me all the clicks on this banner(object id) for this client (client id)"
[11:14:12] <cromeno> A client have many objects(banners) and manny campaigns(A campaign have many objects)
[11:15:14] <cromeno> Hard for me to get out of the rdbms thoughs, so I want to use this project to go mongo
[11:26:32] <cromeno> how do you do then if you have for example a list of products that takes more 16mb in space? For example a shopping website.
[11:28:17] <cromeno> hehe, sorry this noob questions, I will go read the docs, time to go mongo I hope
[12:23:52] <newbieeeee> but it wil think "date.timestamp2" its a word not a refer to a collunm isnt it?
[12:26:06] <newbieeeee> i thought it because i done other thing with that but i was onderring if i was doing something wrong...
[12:29:09] <NeoNmaNDK> zakG: orh okai, i know there are problems on MacPorts on Mac, i don't know about Ubuntu,
[12:30:35] <zakG> actually when i am issueing the command to start mongodb,it shows that the process is started/running but i coudn't find the process in the system monitor
[12:39:12] <kali> it don't know what the system monitor is, I assume it is some kind of ui wrapping ps. anyway, you should look out mongodb logs
[13:04:54] <zakG> NodeX,there's some kind of problem with the mongod.lock file.mongo is gets shutdown automatically.
[13:26:57] <s0urce> Is there any way to get a list of values returned by "find"? I mean instead of "...find(...).toArray() -> {'foo':1,'foo':2,'foo':3}" something like this: {1,2,3}
[13:27:25] <logik-bomb> guys, how do I seed a txt file into mongo? I'm getting SyntaxError: missing ; before statement script_mongo.txt:1
[13:31:21] <logik-bomb> NodeX: https://raw.github.com/gist/3cf13896326aa9c32d29/bbd64d9ecfa71b5c78334657773e5a310b9f7e29/gistfile1.txt
[13:37:23] <Bofu2U> For anyone using the Ruby driver: had issues with it setting the max BSON doc limit to 4 MB when the server is 16?
[14:20:47] <madrigal> hjb_: mongodb docs say If security is configured for a mongod instance, auth is required for a client to access the http interface from another machine
[14:23:28] <hjb_> madrigal: got an explanation: i'm running 2.0.5 and was puzzled by the problem fixed in issue SERVER-2917
[14:39:42] <zenista> hi i want to update a doc... but instead of replacing or modify existing fields i want to append more data to it... ie. say collection of images: {....}
[14:42:06] <devdazed> Hi all, I downloaded the 2.2rc0 to try out the new aggregation framework and i get this "no such cmd: aggregate"
[14:46:05] <bjori> devdazed: yes, you are using the mongo command line shell version 2.2.0-rc0, but you are connecting to your old mongod 2.0.2 server
[14:51:59] <lonnylot> Is there a way to determine the size of a specific object property in a document?
[15:00:09] <NodeX> if you're trying to count the number of array members for example the common approach is to save the count aswell
[15:03:25] <lonnylot> NodeX: I am recording stats that are broken down by year/day/hour. I am trying to determine how many hrs. I can add stats before there is a problem w/ the size of the document.
[15:05:09] <lonnylot> Well we are adding stats as properties to the document. So each yeardayhour will be a different property. I am trying to figure out how many of those can be added to a document before there are problems w/ the document size.
[15:07:06] <lonnylot> Do you know if there are any commands to help figure this out or would it be best to just count the bytes and do the math?
[15:18:17] <kali> lonnylot: why don't you serialize some sample doc and check out the resulting size ?
[15:18:18] <NodeX> if your documents are creeping that high then performance will take a hit as the document goes over the wire
[15:20:14] <lonnylot> NodeX , kali : thanks for the warnings - I will definitely take that in to consideration
[15:23:55] <soumya> i am trying to run some regression tests to see how my query performs with and without indexes.... everytime i run the query it runs faster bacause of data moved to RAM. how do i run a test where the same query runs consistently
[15:26:38] <kali> soumya: if you're running linux, you can try this "free && sync && echo 3 > /proc/sys/vm/drop_caches && free"
[15:53:04] <revnoah> hello, I'm having trouble with the syntax for a mongoose query using a regular expression on multiple fields. I hope someone can point out what's wrong with the statement: app.User.find().or([{ 'firstName': { $regex: re }}, { 'lastName': { $regex: re }}]).sort('title', 1).exec(function(err, people) { /* do stuff */ });
[15:55:00] <revnoah> I have the regex working well enough with a single field, it's the 'or' clause that's giving me problems.
[16:09:52] <dbler> "You may use sort() to return data in order without an index if the data set to be returned is small"
[16:32:30] <W0rmDrink> yeah, I made a query, checked few fields, more or less got what I was looking for
[16:34:36] <dbler> You may use sort() to return data in order without an index if the data set to be returned is small" does it mean that I should NOT use an index if I return little data?
[16:40:19] <soumya> has anyone seen this issue where every consecutive run of a query is faster than the previous one?
[16:54:41] <SNow> hi guys, i want to install mongodb infrastructe on openvz, but I see that even error log tells me that openvz and mongodb seems to be broken, how actually it is?
[17:01:51] <dbler> not this way, I meant redundancy in the bad way: I have for example products belonging to categories, in relational databases I would save the category ID and then do a join to the category table
[17:05:47] <lizzin> how do you increment all documents with a field 'x' by 7? i thought db.stuff.update( , {$inc : {x:7}}) would work. but it fails?
[17:07:16] <mids> first param is the criteria; the query that selects the record(s) you wanna update
[17:12:07] <Vile> dbler: you can handle referential integrity manually (on the software level) or just embed necessary data into documents and do not care about extra data
[17:12:10] <lizzin> mids: so {} appears to work. but it seems to only update the first document. what's need to do it for all?
[17:13:59] <dbler> what do you mean with "on the software level", looking at the category example, would I save then the ID in the database and the name in my source code or what do you mean?
[17:16:49] <slava_> if I want documents to refer to each other, is this something that can be handled on the mongodb side or is it something my app has to enforce?
[17:19:19] <madrigal> lizzin: multi - indicates if all documents matching criteria should be updated rather than just one.
[17:24:13] <dbler> in the first moment I thought MongoDB would be perfect for me (product catalog with about 100 attributes in general, each object has about 35, in OOP a 5 step depp inheritance), but now I have some doubt
[17:31:35] <dbler> slava_: what I think: MongoDB: categoryID = 2; PHP receives this categoryID like that: $catId = getCategoryFromMongo(); $categoryName = $joinInformation['categoryID'][$catId]
[17:37:29] <jrm2k6> Hi guy, i want to deploy a mongodb database on heroku, using the mongohq addon. I m getting this error : Error: database names cannot contain the character '.'. I know it is pretty straightforward but the thing is that the URI for the database is generated by mongohq, using an email address and a password, or, my email contains at least one point :) Any idea? What I am missing?
[18:05:46] <nofxxxx> lizzin, you may write a simple script that imports your CSVs from msaccess (if moving the data was the question)
[18:06:38] <nofxxxx> or, use some dbmanager software to export json directly, should be some one with drivers for that trash
[18:21:51] <madrigal> lizzin: use mdbtools to export your tables from msaccess to csv, there are several "csv to json" options available depending on your config
[18:24:35] <pbakkum> hi all, my shard balancer appears to be stuck in a state where its totally unable to split any chunks
[18:24:55] <pbakkum> in the logs i see the error 'max number of requested split points reached before the end of chunk'
[19:03:12] <lizzin> madrigal: so i've installed mdbtools. but mdb-export's params requires a 'table' what do i use for this field if i do not know anything about the structure of the database?
[19:13:34] <lizzin> madrigal: in msaccess, a table is the overall container for the data right?...similar to a collection in mongodb
[19:14:31] <lizzin> Copy of tblViolationTypes tble_violexplanation tblInspections tblInspectionTypes tblInspectionTypes_old TblInspName tblLocations tblLocationtypes tblViolationCategories tblViolations tblViolationTypes
[19:15:51] <madrigal> lizzin: mdb-tables -d ',' filename.mdb will show you comma delimited table names
[19:17:39] <madrigal> then you can use something like csv2json.py to convert the output: http://bit.ly/OgcBax
[19:18:26] <madrigal> and run the following for each table to convert to json: mdb-export filename.mdb table_name | ./csv2json.py -
[19:19:29] <lizzin> oh, was under the impression that ',' delimiters was the default. http://linux.die.net/man/1/mdb-export. appears to not be though
[19:43:18] <markgm> Can anyone help with the error : doctrine/mongodb dev-master requires ext-mongo * -> no matching package found
[20:11:39] <jrm2k6> I have sonething like that mongodb://my%2Etest@gmail%2Ecom:password@staff%2Emongohq%2Ecom:10056/app6186142
[20:14:53] <madrigal> jrm2k6: yes, '@' symbol must be escaped, try mongodb://my%2Etest%40gmail%2Ecom:password@staff.mongohq.com:10056/app6186142
[20:31:04] <jrm2k6> madrigal: even when i escape everything still the same Error Failed to connect mongodb
[20:44:39] <dstorrs> jrm2k6: I have not used either of those so can't offer much help. but most times, escaping is just \<whatever (e.g., escaping / is \/)
[20:45:04] <dstorrs> sometimes you need to double the \ if there is shell interpolation or etc going on
[20:45:27] <dstorrs> question for the room -- if I rename a table, its indexes are brought along with it, right?
[20:45:30] <jrm2k6> i try to escape everything but each time i escape something the value return by the error is different..
[20:46:21] <dstorrs> it sounds more like a heroku / herokuHQ issue than a mongo issue. I would try their channels.
[20:57:35] <nick_h> why does db.product_haves.find( { 'sale_config.items': { $exists: false } } ).count() always return 0, despite documents in the collection having a "sale_config.items" property?
[21:37:07] <pbakkum> it looks like sharding is stuck in a broken state on my cluster, i keep getting a "max number of requested split points reached" error
[22:05:09] <kida78> hey all - wanted to discuss a few things around writes and performance in mongodb
[22:06:13] <kida78> is there a major difference in performance between my_coll.update({_id: xxxxx }, { entire document }) vs. my_coll.update({id:xxxxxx }, { field 1: value 1, field 2 : value 2}))
[22:06:43] <kida78> we've gotten use to in our code, appending and updating values in an object over time before saving the entire thing in mongodb
[22:35:03] <awestwell> Exception in thread "main" java.lang.NoSuchMethodError: com.mongodb.DB.getOptions()I
[22:39:27] <kida78> awestwell: are you sure you have the right version? is it in your class path correctly?
[22:49:52] <mFacenet> is there an easy way of adding & removing replicaset members if I'm doing an autoscaling group in ec2?
[23:18:19] <hoverbear> Hey all... I've got a database (govtest) with a collection (gov)... When I go into the mongo console I can use db.gov.find() and see {a:'b'}... But using http://www.fpaste.org/vdcP/ I only get 'undefined' when using node.js with the official driver