Log file Viewer

[03:24:01] <sabrehagen> i'm wondering how to design my data to have it perform best in indexes. i have an id field which contains ids from third parties. they take a variety of forms, but are all unique. this will be my primary method of accessing this collection. would indexes perform better if i hashed the value to something of consistent length, rather than using the raw/highly variable data?

[03:25:18] <sabrehagen> e.g., a few ids might be ['a3d2ccaf2s47f2e1a', 'id:33a/2$a1@ask32hh72', 'lkjasdfoialkj83zkl8f32zh782398unk23.99998a8283a.4a831y3333311111112']

[03:36:12] <sabrehagen> also, i want to pick the first document in a group in an aggregation pipeline, rather than just a field of the first document. i'm currently using the $$ROOT variable to pick it out. can this be done better? https://gist.github.com/sabrehagen/836f364a6cd4479e82d25d76dd1e90bf

[03:38:16] <sabrehagen> $$ROOT seems a bit ugly...

[04:03:54] <sabrehagen> a third question. this is my data structure, my aggregation pipeline, and the output. as you can see, the results should be sorted by the modified field, but they come back out of order respective to modified. why is this happening?

[04:05:28] <sabrehagen> joannac?

[06:01:08] <energizer> is it possible to connect to a remote mongo using pymongo through ssh

[06:47:52] <trq> I would assume so. Just make an ssh tunnel.

[07:26:04] <wwwi> hello

[07:26:34] <wwwi> a nosql db has no schema, no foreign keys, no relationships, right?

[07:26:52] <wwwi> doesn't that make it hard to work with it?

[07:28:14] <wwwi> second question, a nosql db is faster than an sql db?

[07:28:59] <wwwi> third question, if you need to scale, pretty much you provision new servers, add something in configuration and you just scaled the db?

[07:29:21] <wwwi> third question, how do you backup a distributed nosql db?

[07:29:57] <wwwi> fourth question ^

[08:12:58] <Dreamcooled> Hi. Is someone around who managed to build the mongo-cxx-driver on windows? I can't get rid of some "undefined reference"- linker errors.

[09:55:41] <onis> Hi guys, in this code: https://gist.github.com/Fen24/a47e0ed30bbf4ca0f98c70e5baabaaf3 is possibile hide the "score" field like i have done with the "id:0" field or i must delete that field only on client-side?

[09:56:22] <onis> yesterday someone told me that is possibile only on client-side but i asking again just to be sure

[10:04:41] <Derick> that was me :)

[10:11:27] <onis> oh yes, seem that we have the same time zone :D, are you still to the same idea? only client-side? yesterday before i left i told that is strange because can be too expansive to do with huge amount of data

[10:12:48] <Derick> Yes, I still think so.

[10:13:19] <Derick> You actively *add* the score with score: {

[10:13:20] <Derick> $meta: "textScore"

[10:13:31] <Derick> so you can't combine that with a 0

[10:13:40] <Derick> I guess you only need to use the score for sorting?

[10:14:13] <Derick> But you're doing a limit(1) - so, what is the issue that you're afraid of?

[10:18:28] <onis> because actually i use that query more than one time then combine the results, so at the end there can be "n json data"

[10:20:15] <onis> but seems that this is the only solution so no problem, i will try your suggestion

[13:48:25] <ams__> I'm trying to sync from mongodb 3.0 to mongodb 3.2: 3.2 is primary, 3.2 is secondary. I have got the error "Unexpected field durableOpTime in UpdateInfoArgs" and the sync has stopped

[13:48:27] <ams__> Any idea why?

[13:49:59] <ams__> There are literally 0 results on google for that string, which is fun :)

[14:29:04] <bros> My production server randomly spikes (and holds) 700% CPU usage for about 15 second bursts randomly. How can I figure out what is causing this?

[16:48:21] <wreckram> hi

[18:43:17] <naf> ams__

[18:43:27] <naf> try searching just for "UpdateInfoArgs" and paging thru the src

[18:43:35] <naf> https://github.com/mongodb/mongo/blob/master/src/mongo/db/repl/old_update_position_args.cpp

[19:05:42] <angular_mike> can I comfortably have a database with about 2 write operations per 15 min of ~100 documents(~40 fields), 2 read operations on 101 documents per 15 min on 1-vcpu machine with 0.6 GB ram?

[19:06:06] <StephenLynx> not comfortably.

[19:06:14] <StephenLynx> hm

[19:06:16] <StephenLynx> maybe?

[19:06:42] <StephenLynx> I remember running a server with 512mb with both the db and the application on the same server.

[19:06:44] <StephenLynx> and it ran.

[19:07:09] <StephenLynx> but anything short of 1gb I wouldn't call comfortable.

[19:10:10] <LouisT> i had it running with more writes than that with about 512MB (256MB swap) with other random things on the same VPS for a while, it wasn't GREAT but it worked ok while i was developing things

[19:19:51] <angular_mike> Oh my goD! someone actually responded

[19:20:01] <angular_mike> I did not even realise

[19:20:23] <angular_mike> LouisT: how not great?

[19:20:36] <StephenLynx> kek

[19:21:02] <angular_mike> LouisT: I'm also planning to run a minimal http server for low-rate requests on the same machine

[19:21:23] <LouisT> angular_mike: things ran a little slower than i planned, it was usable but i wouldn't want to run anything production like that ever

[19:21:55] <angular_mike> I mean, did you have to take pauses till newest batch of data gets in for new iteration, etc?

[19:22:18] <angular_mike> is just a prototype for fun and experiments

[19:23:29] <LouisT> nope, was able to read/write fine for the most part without having to queue anything, it just made other things a bit sluggish on the server

[23:10:11] <maumercado> Hi, Im having an issue with an index

[23:11:37] <maumercado> I created a ttl index as a background, and stopped it when it was around 75% completion, and now I need to rebuild it, but I cannot drop it

[23:11:42] <maumercado> or recreate

[23:12:12] <maumercado> message error while attempting to drop is cannot find index, while error message when recreating is index already exists :/

[23:12:38] <maumercado> using mongo 3.0.8

[23:13:14] <cheeser> and you can't drop it?