Log file Viewer
#mongodb logs for Saturday the 16th of May, 2015
[00:05:36] <fructose_> StephenLynx: I think I'd have to omit the _id then. Is there a value for _id I can pass that would be the equivalent of omitting it?
[00:07:53] <fructose_> StephenLynx: I'm trying not write two different things to do insert and update. That seems to be the purpose of upsert. So if I'm passed an ID and I don't know if it exists or not already, what do you suggest?
[00:09:49] <StephenLynx> when I used upsert I didn't had that problem because of that, I didn't relied on _id for the update, but on a different field.
[00:10:37] <StephenLynx> you could use application code to add the _id field on the query and projection blocks depending if the _id was informed.
[00:11:23] <StephenLynx> or try and see if you pass the _id as null will make mongo realize what you are trying to do and just create a new _id on its own.
[00:21:14] <fructose_> Can anyone tell me why I'm getting no error, doc is 0, and nothing is being inserted here: http://pastebin.com/id7KWcsP
[00:24:00] <fructose_> StephenLynx: As long as they're not to hold me hostage before you tell me what's wrong :)
[00:25:09] <StephenLynx> don't use express, don't use mongo. use just the runtime environment and the node.js driver.
[00:25:50] <StephenLynx> when you are inside a callback, use conditions and write nothing but a single line inside of each conditions, each line proceeding the logic on a separate function.
[00:31:08] <StephenLynx> that means I couldn't understand jack because of the ODM that requires the knowledge of a sub community and the bad formatting.
[00:33:10] <fructose_> Are you referring to monk? Because I don't see how anything else is even relevant
[00:34:24] <fructose_> What are you suggesting I use instead? You said "don't use mongo". Did you mean monk?
[00:37:10] <st34lth> StephenLynx: curious case, right, I create a user and can log in from command line, but none of the client will auth using user name and password, that I can verify works in the cli
[00:37:54] <StephenLynx> froctose_ mongoose is cancer, almost any issue with people using mongo with node/io is because of mongo being bad.
[00:38:05] <st34lth> in the log I can clearly tell 2015-05-15T20:29:30.121-0400 I ACCESS [conn2] Failed to authenticate dbadmin@admin with mechanism MONGODB-CR: AuthenticationFailed MONGODB-CR credentials missing in the user document
[00:38:44] <StephenLynx> IMO any tool that does anything you can already do, but does different is bad.
[00:39:46] <StephenLynx> I managed to connect to instances with auth on by simply formatting the connect string.
[00:41:01] <fructose_> StephenLynx: Is there a tutorial you'd recommend for the native node.js driver?
[00:42:09] <st34lth> I'm fine with the cli, but I need to show the data to other folks, and I automatically taught of a client, postgres, mysql, etc all have clients like that, so i looked into robomongo and mongohub.
[00:45:53] <st34lth> I'm not even creating data yet at the application level because I wasn't sure if application will also require MONGODB-CR
[00:47:51] <st34lth> alright, I was following the docs, http://docs.mongodb.org/manual/tutorial/enable-authentication-without-bypass/
[00:53:03] <st34lth> I took your advice, connected on first try without creds in mongohub and robomongo StephenLynx
[00:53:44] <st34lth> I went on a tangent trying to fix this auth thing because I thought that will be a standard. next time, I just stick to book. lol
[03:45:57] <erobit> does anyone know the chance of collision for object uuids in mongoose driver across collections / within same db?
[03:48:39] <erobit> I'm using mongoose-version for versioning my schemas and was wondering if its safe to apply it as a plugin to multiple schemas - it uses refId -> objectid as the key to store versions of a document as it changes
[03:50:53] <Boomtime> objectid, per host that is generating them, can generate 16 million unique values per second before collision becomes imminent
[03:51:32] <Boomtime> that may extend to individual process (not just per host) depending on the driver
[03:52:26] <erobit> I think I'll be alright, cause each schema that is being versioned is its own collection as well - so those ids should be unique.. ok I shouldn't really worry about it then
[03:53:03] <erobit> it's not like I'll be inserting more than 20 records per second - not an app that needs to support thousands of users either. awesome thnx again
[14:16:54] <nalum> hello all, can anyone tell me why members of a replica set would be stuck in the stateStr STARTUP2?
[16:03:36] <Zerith> hey. I'm using pymongo and a local MongoDB server. I'm iteratively calling insert_one() and find() on many collections, this worked for some entries, but now I can see that my insert_one() calls aren't inserting any data (and I can't see anything in the mongodb server logs either) does anyone know what might be the problem?
[16:39:52] <Zerith> I want to show some people the database and I don't think the CLI will be that convenient
[16:40:28] <GothAlice> I have that as a shortcut, with replacements. (I type *fo<tab>, the name of the collection, tab, and paste in an ID.) I have a number of shortcuts. (TextExpander) I prefer to use ipython or bpython for their syntax-highlighting; they're Python, though, so pymongo.
[17:35:54] <RedWraith> Is there a way to search a document and modify every found element in some way that uses the original value as a parameter?
[17:44:46] <StephenLynx> depending on what you need, there are even update operators, like the multiply one.
[17:46:58] <StephenLynx> there is one thing that bummed me out with $out on aggregate, if the collection exists, it will flush the existing values.
[17:51:27] <RedWraith> StephenLynx, could you give me an example of how you might use $fieldname for an update? I tried it and I think I did it incorrectly - in CoffeeScript: Circles.update {}, {data: _.shuffle $data}
[17:55:07] <RedWraith> StephenLynx, sorry, I'll convert it to JavaScript then: Circles.update({}, {data: _.shuffle($data)});
[17:56:16] <StephenLynx> and there one thing, you can't just give mongo your functions to run on your application.
[17:57:51] <StephenLynx> which was what I ended up doing on my application when I found an issue like this.
[17:58:24] <StephenLynx> I couldn't figure out map reduce so I just went with a regular find and iterated the documents.
[17:58:51] <mike_edmr> i usually prefer small updates to big ones, the big ones seem to cause bigger interruptions for other concurrent work
[17:59:11] <mike_edmr> in fact i have a lot of code that "trickles" in updates to mongo to avoid overloading it
[18:01:05] <mike_edmr> large updates make the database unresponsive to the massive amount of reads happening
[23:54:59] <android6011> I'm using mongojs to extract some data from postgresql and insert to a mongodb. I am attaching a a mongojs.ObjectId() to each object before sending. I keep track on the app side of how many diff ones have been seen, for testing I only get 1, but when I look in mongo I have several 'distinct' values for that field