Log file Viewer
#mongodb logs for Monday the 7th of September, 2015
[00:02:37] <damoncasale> StephenLynx: Reading over what I can find on MongoDB performance vs. Mongoose performance. It seems like Mongoose can still work, but one would need to be careful to use it correctly.
[00:03:25] <damoncasale> For instance, one of the slowdowns appears to be the validation prior to writing out a record. Mongoose provides a bare query method, tho.
[00:07:23] <StephenLynx> Mongoose assigns each of your schemas an id virtual getter by default which returns the documents _id field cast to a string, or in the case of ObjectIds, its hexString.
[00:07:52] <StephenLynx> in the last year, every single issue people had with node and mongo was related somehow to mongoose.
[00:08:16] <StephenLynx> and since you just started learning mongo, I assume you are not bound by some legacy system using this cancer.
[00:29:23] <mprelude> * (cl-scram-sha1:base64-encode (cl-scram-sha1:gen-client-initial-message :username "aurum" :nonce (cl-scram-sha1:gen-client-nonce)))
[00:37:52] <mprelude> StephenLynx: Decided the only way forward is to implement SCRAM-SHA1 in lisp. :D
[00:38:35] <mprelude> Hoping I can have that done, and the mongo driver updated to use it, this week.
[06:01:05] <pEYEd> how would I query all of the usernames from this collection? http://paste.ee/p/o9xGf it's inside the subarray of 'data' and I can't get to it. :(
[09:01:18] <jebis> Hello, is there a way in php to make a ordered insert of multiple documents? u know i want to have them in a ordered manner http://pastebin.com/EmwSsztD i dont know it is allways mixed even when i fetch them with sort "create_date" : -1
[09:21:58] <jebis> coudenysj this natural works but why does not work a normall sort by create date ? i mean is mongo so fast that it creates 2 same dates ?
[09:24:05] <coudenysj> the first part of the id is the time, but there are other values needed (mostly to solve multi server problems), but you can insert multiple docs in a given msec
[09:33:02] <jebis> coudenysj and $natural uses that part of the id ? or how it works if i can ask :P
[11:07:36] <agis-> Isn't something wrong with my 2.4 server logs? https://gist.github.skroutz.gr/agis/212d40ee53e5ac7e3006
[11:11:32] <mprelude> Hi, sorry if this seems like a noob question. I'm updating one of the mongo drivers to use SCRAM-SHA1. I was wondering the easiest way to get into a communication with the server so I can try throwing requests at it and look at the responses.
[11:12:43] <mprelude> I just want a raw TCP discussion with the server, I guess, so I can try things.
[12:30:28] <mprelude> cheeser or Derick (or anyone else): If you can help me get pointed in the right direction to do a mongo auth over telnet or similar, that'd be great. I can't seem to find any tcp-level documentation.
[13:29:10] <nerder> which one of this 2 solution can be better in terms of performance? https://gist.github.com/nerder/ee5edac69f8306f11bfd
[13:30:03] <nerder> they produce actually the same result, in the first case i use ArrayList because $in wants a List
[13:55:36] <corentin> I've seen about {emtpycapped: "collection"} but it seems to only be a test command
[14:06:21] <mprelude> Derick: The main thing I'm struggling with is just how to format my requests to the server. Usually I do this kind of stuff over telnet so that I understand what i'm implementing, then write them in code.
[14:06:56] <Derick> mprelude: this is what PHP does: https://github.com/mongodb/mongo-php-driver/pull/730/files
[14:08:52] <Derick> but not sure if you can do it with telnet, as it requires binary info to talk to mongodb
[14:09:41] <Derick> mprelude: and https://github.com/mongodb/specifications/tree/master/source/auth is our spec for it
[14:12:27] <Derick> https://github.com/mongodb/specifications/blob/master/source/auth/auth.rst#scram-sha-1 to be about scram-sha-1
[14:15:08] <crispy_beef> Hi. Wondered if somebody here could shed some light on why we are getting the error message "no primary server available" now and then when connecting to our replica set?
[15:56:23] <corentin> Is there some way to automatically delete a collection after some timeout has expired?
[16:02:53] <corentin> cheeser: thanks. Is there no way to automatically delete a collection when it's empty?
[16:13:20] <corentin> cheeser: I think it's mostly useful when using mongodb for web stuff, such as django or so
[16:19:59] <corentin> cheeser: well in our case we have limited hard disk space, so we have to use capped collections and rotate the documents in our collection. This means when we display a document to a user, the document might have expired if/when the user request more data about this document
[16:20:21] <corentin> our solution was to copy the current collection to another collection with a name containing the user session ID
[16:47:53] <cheeser> you'd be better off to not use dynamically created collections and just ttl away the data you don't want
[16:59:01] <StephenLynx> why dont you just use ... options,callback); instead of using a dummy callback?
[17:00:15] <synthmeat> if you know of a more convenient workaround that, i'd be glad to refactor my stuff, since there's a lot of this ugly stuff around
[17:00:53] <synthmeat> no, there's in document alread stuff under facebook., and if i fill up facebook object and pass it as update, what was there disappears
[17:03:08] <synthmeat> ok. how about this. i fo refacctor my stuff to not use mongoose and then i'll find this channel much more useful? :)
[17:07:16] <StephenLynx> http://stackoverflow.com/questions/15724272/what-is-the-difference-between-id-and-id-in-mongoose
[17:07:19] <StephenLynx> Mongoose assigns each of your schemas an id virtual getter by default which returns the documents _id field cast to a string, or in the case of ObjectIds, its hexString.
[17:08:30] <StephenLynx> and no one from the people that I have ever seen here giving advice on anything uses it.
[17:38:35] <synthmeat> what was expected was new document if one was found and updated, and nothing if no document was found
[17:52:59] <synthmeat> this is the code producing that error https://gist.github.com/synthmeat/72ec11e75c06dd7e6fe9
[18:02:20] <synthmeat> StephenLynx: i still have issue with that pasted code in that it returns response, creates document
[18:20:34] <flok420> in an aggregate I want to group records by their timestamp per hour. timestamp in seconds so I thought I just do ts - (ts % 3600). Am I right that that should be written like this: $group: { '_id': { name : "$name", ts : { $subtract : [ "$ts", { $mod : [ "$ts", 3600 ] } ] } }
[18:20:56] <flok420> because mongodb shell says "SyntaxError: Unexpected string" (with no hints where this faultive string is)
[18:22:12] <StephenLynx> so it will always be different even if the part relating to time is the same
[18:23:07] <flok420> my bad I meant that i want to group on the name-field together with the timestamp. yes, storing names together with epoch-times (seconds between now and 1970)
[18:23:55] <synthmeat> StephenLynx: fyi, yes, it was me. query value was undefined. not sure why it acted like it found something and created still, but that was it
[18:24:39] <StephenLynx> but consider using date objects, that allows you to perform operations on the date at a database level.
[18:25:31] <StephenLynx> it can't tell you very well where the error is because of how json is formatted.
[18:27:45] <flok420> oh hang on it indeed has nothing to do with that maths I added. if I remove it (and only group on name) it still complains about a string! ok thanks sofar, I now have something to work with
[19:31:38] <najah> Hi all, I'm a young sysadmin and I use mongodb with docker. I would like to know things about journal in mongodb. I would like to know what are pros and cons about journal in mongodb for a little application. It is working like RDBMS ? I mean... Can we start to a special point in the journal before error humain ? Thanks very much
[19:33:42] <treta> I'm getting this message all the time, it's not fun anymore "Caused by: com.mongodb.MongoTimeoutException: Timed out while waiting for a server that matches AnyServerSelector{} after 1000 ms"
[20:10:40] <frenchiie> might someone be able to help me with this question? http://stackoverflow.com/questions/32444803/pymongo-parameters-of-function-find-one-and-update
[20:56:05] <sqram> for instance, if i have a document like {id:4, children: [4,3] } ... can i update the values of children in a single update command? or do i have to findOne({id:4}), alter the chuldren array, and then update
[21:01:25] <sqram> now going a step futher..say i want to update multiple documents - determined by their unique ids. is it possible?
[21:02:29] <sqram> for instance, toupdate = [3,7,8,5] <-- these would be id fields of different documents. must i run a loop to update each one, or can a single update call do it?
[21:02:51] <sqram> i di dsee multi: true in the docs, but seems like it would have to be a field value shared among the documents. these are unique
[21:17:46] <mprelude> Derick: When I'm creating HMACs, what format am I looking to get the digest out in?
[21:18:21] <mprelude> Derick: Obviously, hashing a string of a hex num will be entirely different to hashing the same num in decimal or binary.
[21:32:20] <mprelude> Would be nice if the example showed all the intermediate values so we had something to test against.
[21:38:20] <mprelude> Like, the proof in the spec, MC2T8BvbmWRckDw8oWl5IVghwCY= ... appears to base64_decode to something utterly useless.
[21:39:12] <mprelude> I'm not sure what it represents, I don't know how to test whether my implementation is right :(
[21:48:05] <mprelude> http://stackoverflow.com/questions/29298346/xmpp-sasl-scram-sha1-authentication I've been working off this answer and the rfc
[21:49:29] <mprelude> Literally the only part I'm really stuck on is what format these hashes should be on. I have to do logical XORs and re-hashes, which could be on binary, decimal, octets, hex idk.
[21:51:11] <Derick> mprelude: https://github.com/mongodb/mongo-c-driver/blob/master/src/mongoc/mongoc-scram.c#L395 is step 2 in the c driver