PMXBOT Log file Viewer

Help | Karma | Search:

#mongodb logs for Wednesday the 22nd of July, 2015

(Back to #mongodb overview) (Back to channel listing) (Animate logs)
[00:12:48] <Ned_> We have a replicaset in MMS that doesn't have automation. We checked the box to enable automation, and we just get a blank page (and nothing happens). Anyone know what might be up there?
[03:04:56] <edward103> hello guys
[03:05:12] <edward103> i need help
[03:05:49] <edward103> mongodb c# driver
[03:19:37] <Boomtime> hi edward103, we'd like to help you but you need to hang around for more than 2 minutes to give people a chance to respond :-(
[03:25:56] <Ned_> Are there any MMS experts here?
[03:28:23] <joannac> Ned_: You should just ask your question.
[03:29:11] <Ned_> 12:09 < Ned_> We have a replicaset in MMS that doesn't have automation. We checked the box to enable automation, and we just get a blank page (and nothing happens)
[03:29:15] <Ned_> 12:10 < Ned_> We have a replicaset in MMS that doesn't have automation. We checked the box to enable automation, and we just get a blank page (and nothing happens). Anyone know what might be up there?
[03:29:20] <Ned_> oops, sorry
[03:29:27] <Ned_> copy/paste fail
[03:29:34] <joannac> Open a MMSSUPPORT ticket?
[03:29:40] <Ned_> we've done that
[03:29:48] <joannac> what's the ticket number?
[03:30:05] <waa_> joannac: https://jira.mongodb.org/browse/MMSSUPPORT-7193
[03:30:10] <Ned_> joannac: ^^ that's the ticket
[03:30:35] <joannac> Thanks. You're in the queue
[03:34:03] <Ned_> okay, thanks ...
[03:42:33] <joannac> Ned_: If it were a simple question I could answer, but it's going to require investigation :)
[03:42:58] <Ned_> joannac: yeah, we fiddled around a bunch before submitting the ticket, but we're pretty confused about it too now :-9
[03:43:01] <Ned_> :-(
[04:24:56] <_sillymarkets> Hey, I have a question regarding "bulk_write" & the ruby driver. The mongodb/ruby-driver page on Git with examples doesn't seem to work
[04:25:39] <_sillymarkets> I get undefined method "initialize_ordered_bulk_op"
[04:26:06] <_sillymarkets> client[:collection].initialize_ordered_bulk_op gives me undefined method
[04:31:21] <joannac> _sillymarkets: what version?
[04:32:02] <_sillymarkets> mongodb 2.6.3
[04:32:29] <joannac> what version of the driver
[04:33:50] <_sillymarkets> 2.0.6
[04:34:09] <_sillymarkets> "gem install mongo"
[04:34:22] <_sillymarkets> I found for the aggregate command, I had to use the following:
[04:34:34] <_sillymarkets> client[:collection].find.aggregate(...)
[04:35:08] <_sillymarkets> as opposed to what the docs or examples on github state client.collection.aggregate
[04:35:34] <_sillymarkets> to trigger the methods under CollectionView
[04:37:14] <joannac> _sillymarkets: I suggest you file a ticket in the RUBY project
[04:37:47] <_sillymarkets> okay
[04:37:49] <joannac> I haven't found the page(s) you're referring to - everything I've looked at has "coll.find.aggregate"
[04:38:53] <_sillymarkets> https://github.com/mongodb/mongo-ruby-driver/wiki/Aggregation-Framework-Examples
[04:39:52] <_sillymarkets> I figured out aggregation though. I'm trying to use their Bulk_Write examples, but they don't seem to go for me.
[04:46:59] <_sillymarkets> happen to know any links with newer bulk_write examples in ruby?
[04:57:00] <_sillymarkets> found some
[05:01:20] <_sillymarkets> hm, how to bulk_write {:insert_one => array_of_hashes }. If I iterate through that array to get each hash inserted, its not really a bulk_write. It's the same as .insert_one(:params)
[05:02:37] <_sillymarkets> is what I'm attempting even possible?
[05:27:52] <_sillymarkets> hm, so i figured it out
[05:28:37] <_sillymarkets> but .bulk_write([ {:delete_many => array}) still seems slow and one by one to the db
[05:41:38] <mbuf> is there a way to test cloud.mongodb.com API keys with curl?
[05:57:00] <mbuf> should /data/db be located at /, or should it be relative to a mongo installation directory?
[06:26:55] <Boomtime> mbuf: are you asking about the actual path "/data/db" (this is a full path as is), or do you asking about best practices for determing what to use for a dbpath?
[06:27:29] <Boomtime> wow.. lots of mistakes in that sentence, please adjust as appropriate
[06:34:56] <mbuf> Boomtime, best practices; running mongodb inside a Docker container
[07:16:37] <Boomtime> mbuf: sorry for the delay, i think it's a very subjective question - depending on certain surrounding conditions you may be able to make the answer more objective though
[07:16:39] <mbuf> Boomtime, never mind, used /etc/mms/data/db
[07:17:22] <mbuf> does anyone know how I can check that the mms-agent is running on a mongodb instance?
[07:18:53] <Boomtime> ps -aux | grep mms
[07:18:55] <Boomtime> ?
[07:24:12] <mbuf> Boomtime, don't see an entry. How else can I check its status? Is there some documentation I can refer?
[07:32:08] <mbuf> "sudo service mongodb-mms-backup-agent status", says "mongodb-mms-backup-agent: unrecognized service"
[07:32:58] <mbuf> "sudo start mongodb-mms-automation-agent", says start: Job is already running: mongodb-mms-automation-agent
[07:33:18] <mbuf> any other way by which I can verify the same?
[07:35:11] <Boomtime> what does the service status of the autom agent say? -> sudo service mongodb-mms-automation-agent status
[07:35:29] <Boomtime> (it should be the same as the 'start' command)
[07:35:55] <Boomtime> so do a sudo ps -aux | grep mms
[07:38:27] <mbuf> Boomtime, sudo ps -aux | grep mms gives "guest 18007 0.0 0.0 10460 936 pts/1 S+ 07:35 0:00 grep --color=auto mms"
[07:39:02] <mbuf> Boomtime, "sudo service mongodb-mms-automation-agent status" says it is running though
[07:39:29] <Boomtime> huh.. i thought that mms was in the process name, but maybe i'm mistaken
[07:41:50] <mbuf> I don't see anything in the dashboard either. Will try to change API key and re-start the MMS agent and see
[07:42:17] <Boomtime> see if it responds to stop first
[07:42:30] <Boomtime> and yeah, i'm not sure what is going on
[07:43:14] <mbuf> Boomtime, yes, it stops
[07:50:18] <mbuf> Boomtime, Eureka! the key was incorrect. I see some logs, but, no graphs or anything
[07:55:52] <Boomtime> mbuf: good to hear, though i'm still confused why it doesn't show up in ps
[08:34:53] <mbuf> are there any screenshots on what to expect in the mongodb CloudManager on how it displays the data received by the MMS agents?
[08:57:48] <mbuf> how do I remove the monitoring and backup agent that got installed from the UI? I started to configure it, but, it automatically installed it
[09:13:46] <KekSi> how do i allow a user with the role userAdminAnyDatabase on admin to access config?
[09:14:28] <KekSi> i get "errmsg" : "not authorized on config to execute command { dbstats: true }" , "code" : 13
[09:26:15] <mbuf> the automation agent and monitoring agent are two different MMS agents?
[09:41:36] <mbuf> is there a way to configure the automation agent to also be used for monitoring and backup? I deploy the agent using Ansible
[10:06:58] <mbuf> the MMS agent must not be run on the same Docker container where the mongod server is running?
[10:07:16] <mbuf> it can be run on the host, instead?
[10:28:13] <KekSi> btw: how do i stop logging this sort of thing: 2015-07-22T10:24:25.875+0000 I SHARDING [Balancer] distributed lock 'balancer/fb58f165dba3:27017:1437468423:1804289383' acquired, ts : 55af6f59721e883cef81ae1e
[11:09:12] <iszak__> I want to get the sum of the occurrence of the first element of a sub-document, I don't think $pop is what I want, nor $first, I tried using .0. but this just gives me an empty array back
[11:39:34] <louie_louiie> Morning Mongo
[11:39:41] <louie_louiie> http://thehackernews.com/2015/07/MongoDB-Database-hacking-tool.html
[11:44:57] <arussel> is there a way to resume a mongorestore ?
[11:46:40] <KekSi> louie_louiie: thats simply because of the default that doesn't need authorization
[11:53:24] <louie_louiie> @KekSi, true. furthermore the article states some versions are missing the bind_ip 127.0.0.1 in the .conf leaving default users vulnerable to remote access
[12:20:13] <KekSi> yeah but that's mostly because people want to access it from somewhere other than 127.0.0.1 and just take that line out (was like that at my old job aswell)
[12:20:35] <KekSi> they just had it running on a rented vserver like that 8(
[12:22:32] <StephenLynx> lelel
[12:24:26] <mydigitalself> hi folks, does anyone know anything about WiredTiger cursor counts?
[12:24:27] <mydigitalself> db.serverStatus().wiredTiger.session
[12:24:32] <mydigitalself> just goes up and up and up
[12:24:35] <mydigitalself> which is mildly worrying
[13:19:54] <lmatteis> guys how do i sort on a NumberLong field?
[13:20:03] <lmatteis> .sort({start: 1}) doesn't seem to work
[13:45:01] <xissburg_> http://cryto.net/~joepie91/blog/2015/07/19/why-you-should-never-ever-ever-use-mongodb/
[13:45:03] <deathanchor> really?
[13:45:29] <StephenLynx> lel
[13:45:37] <StephenLynx> >... is a nightmare to scale and maintain
[13:45:41] <StephenLynx> >opinions
[13:46:22] <StephenLynx> >... forces the poor habit of implicit schemas in nearly all usecases
[13:46:26] <StephenLynx> >more opinions
[13:46:30] <Zelest> they explain 32bit systems and 2.4.x? o_O
[13:46:41] <StephenLynx> that post is crap materialized.
[13:46:50] <deathanchor> StephenLynx: agreed
[13:47:05] <xissburg_> fanboys?
[13:47:09] <deathanchor> clearly a SQL-minded person
[13:47:22] <deathanchor> xissburg_: you are in a mongodb chat channel
[13:47:42] <deathanchor> clearly many of us use it in production all the time.
[13:47:44] <xissburg_> well, I used mongo db a little and found it quite fun
[13:47:51] <Zelest> xissburg_, I agree to be a fanboy when you agree to be a hater. :)
[13:48:08] <Zelest> then i'm not fanboy :)
[13:48:11] <Zelest> no*
[13:48:26] <xissburg_> I have just heard really bad things about mongo db over the years
[13:48:34] <StephenLynx> they are mostly wrong
[13:48:41] <StephenLynx> try it for yourself.
[13:48:47] <StephenLynx> and draw your own conclusions.
[13:48:53] <StephenLynx> and don't expect it to be a rdb.
[13:48:54] <xissburg_> Nothing better than that
[13:49:05] <StephenLynx> you are not in kansas anymore.
[13:49:19] <xissburg_> rdbs feel so antiquated
[13:49:22] <deathanchor> xissburg_: bad things with mongodb is usually due to user/admin/dev errors with implementation
[13:49:27] <StephenLynx> they have their use.
[13:49:50] <deathanchor> yeah rdbs are good for some scope
[13:49:53] <StephenLynx> sometimes you need ACID, you need relational integrity, you need joins, you need transactions.
[13:49:53] <deathanchor> same with mongo
[13:49:58] <StephenLynx> but sometimes you don't.
[13:50:04] <deathanchor> StephenLynx: agreed
[13:50:21] <deathanchor> use the tools best suited for the problem at hand.
[13:50:33] <xissburg_> ^
[13:50:53] <StephenLynx> I can guarantee you that knowing how to work with mongo is better than not knowing how to work with mongo.
[13:51:00] <xissburg_> we often see people forcing stuff to behave in ways they were not designed to
[13:51:03] <StephenLynx> yes.
[13:51:07] <StephenLynx> people do that a lot of mongo
[13:51:08] <deathanchor> StephenLynx: same can be said about any RDB.
[13:51:10] <StephenLynx> and then go rant in the internet.
[13:51:21] <StephenLynx> a lot with*
[13:51:23] <xissburg_> for example, ruby motion to write iOS apps using ruby -________-
[13:51:28] <StephenLynx> lelelel
[13:51:37] <xissburg_> (not sure that is still a thing)
[13:52:27] <StephenLynx> exclusively front-end devs usually are the kind of people that shouldn't be allowed near a computer.
[13:52:34] <StephenLynx> let alone be trusted developing software.
[13:52:47] <StephenLynx> so I don't get surprised anymore by their shenanigans.
[13:52:52] <xissburg_> heh
[13:53:21] <StephenLynx> with mongo, I suggest you rely on the official documentation.
[13:53:31] <StephenLynx> and don't use it with anything on top right from the start.
[13:54:23] <xissburg> anything on top? like what?
[13:54:35] <StephenLynx> like canc- mongoose or anything like that.
[13:54:41] <deathanchor> StephenLynx: mongouniversity M102 lessons really helped me wrap my head around it.
[13:55:08] <xissburg> uh raw mongo db is so easy
[13:55:09] <deathanchor> it's pure mongodb lessons.
[13:55:33] <deathanchor> xissburg: you'd be surprised how much you can do with some js script and mongo
[13:55:56] <xissburg> probably
[14:13:20] <symbol> Is a cursor always returned in the shell even if there's less than 20 documents?
[14:15:54] <deathanchor> if you use find, yes
[14:16:04] <deathanchor> findOne returns a doc
[14:18:50] <symbol> Cool - just wasn't sure if that counted as a cursor.
[14:20:21] <tgreena> Hey guys, I'm fairly new to Mongo (I'm old school MySQL/Post) and I have a question about the best way to set up my database.
[14:21:48] <deathanchor> tgreena: ask away.
[14:22:09] <deathanchor> we usually skip the nicities and just ask questions
[14:22:19] <tgreena> Basically, my question is this: I have logs that I need to input into my MongoDB, these logs contain thousands of rows each. I have currently one 'table' in my database, ( I think it is called a collection in Mongo) and inside that collection, I have just each individual row (Mongo calls them documents I believe)
[14:22:27] <tgreena> I was typing it :D
[14:22:31] <endstille> hey may i ask a mongodb-related queston about rsyslog here?
[14:23:01] <tgreena> Is it best to structure it like this, were in the collection, each row in my log, is a document in my collection
[14:23:16] <tgreena> or should I make a single collection have a subcollection of all the rows
[14:23:27] <deathanchor> tgreena: I would say yes, and if you can split out the info you want to query on into sep fields.
[14:23:49] <tgreena> I can easily split it out into separate fields, in fact I already have
[14:24:08] <tgreena> I have it currently where each row in my logs are stored into a single row (document) in my table
[14:24:13] <deathanchor> tgreena: then you are good, just setup indexes on those things
[14:24:25] <deathanchor> endstille: yes you can ask
[14:24:42] <tgreena> Is it going to present a problem down the road, when I inevitably get to a million, or a billion rows
[14:24:59] <deathanchor> tgreena: TTL index
[14:25:48] <deathanchor> tgreena: or capped collections
[14:25:51] <tgreena> okay interesting, I will have to take a look at TTL index
[14:26:06] <deathanchor> tgreena: or rotate the collection out every day
[14:26:17] <deathanchor> so many solutions
[14:26:21] <endstille> thanks :)
[14:26:21] <endstille> i've setup a RSysLog Server, running on Debain Jessie
[14:26:21] <endstille> it ran for some time, using mysql, now i added MongoDB because it think it will be much faster
[14:26:21] <endstille> it works fine but i have this one issue: i cannot search "freetext" in my mongoDB. if i search for a specific host Host or something, it works flawlessly, if i search just anything, i can see in HTOP, that at least on CPU jumps to 100% for mongo, until the PHP-Timeout is reached
[14:26:51] <tgreena> deathanchor: I can't rotate the collection out every day since we are looking to archive all these logs so our clients can review the past
[14:27:41] <deathanchor> tgreena: I meant you can rotate them daily so you can archive or move collections overtime easily by time
[14:28:03] <deathanchor> http://docs.mongodb.org/manual/reference/method/db.collection.renameCollection/
[14:29:22] <deathanchor> endstille: I don't quite understand your problem.
[14:29:29] <tgreena> Hmm, i'll have to read up because that is a little confusing to me, let me take a look at that doc you sent
[14:29:39] <endstille> deathanchor: i cannot use the search
[14:29:48] <endstille> if i use my mongodb
[14:29:58] <deathanchor> you mean db.coll.find()?
[14:30:26] <endstille> no i don't work manually in mongo, (yet :) ) i just use rsyslog
[14:31:15] <endstille> http://www.rsyslog.com/ and check it via loganalyzer http://loganalyzer.adiscon.com/
[14:31:56] <deathanchor> sorry, I can only help with direct mongo questions. I don't have time to learn new tools.
[14:33:29] <endstille> no Problem! maybe you can guide me, try the manual search once :)
[14:33:47] <endstille> what is the syntax for db.coll.fin()?
[14:34:20] <endstille> i never worked on any Database, except some simple things using phpmyadmin ;) i'm system administrator
[14:34:33] <deathanchor> use db; then db.collectionname.find({ field : "valueyouwant" })
[14:34:47] <deathanchor> you should do the mongouniversity M102
[14:34:59] <endstille> great, thanks :) i'll try right now!
[14:36:29] <bodisiw> hi all, i have a text index on my field, but for some reason one term in a document doesn't match the document, while the others do... is there any special gotcha i should look for?
[14:36:37] <bodisiw> here's a gist demonstrating my issue https://gist.github.com/noahadler/e04bbacb4b4bab5ee5cb
[14:41:53] <lobmys> Did you add that word after the initial index?
[14:43:31] <bodisiw> lobmys, sorry can you explain? do i need to do something special if i update that field on an existing document?
[14:44:49] <lobmys> bodisiw: I haven't used $search much but I do know that it requires the text index on all the values.
[14:45:12] <lobmys> I'm slightly guessing that if you added the word after the index then it's not searchable.
[14:45:19] <lobmys> I'm experimenting real quick
[14:47:00] <bodisiw> i do ensureIndex({searchText: 'text'}) on the collection... then i figured any documents added or updated after that should get into the index
[14:49:20] <lobmys> Gah - you nerd sniped me.
[14:53:59] <bodisiw> well, i tried again with fresh data and got the same issue :-/
[14:59:16] <lobmys> That is peculiar
[15:02:21] <bodisiw> lobmys, i agree :-(
[15:03:15] <bodisiw> maybe i should just use regex... i don't get any lemma/stem help since it's unsupported language anyway
[15:04:08] <bodisiw> still concerning to see this unexpected behavior...
[15:05:46] <lobmys> I suspect there's a nuance here that's being overlooked.
[15:07:12] <bodisiw> that's fair... i'd like to know! maybe it's been fixed in a newer version (i'm running 2.6.7)
[15:08:18] <lobmys> Heh, as would I. I'll keep poking around to see what I can find.
[15:41:56] <lobmys> bodisiw: I'd suggest asking again - I'm sure someone here with more experience will be able to help you out. I'm also selfish and really want to know what's going on.
[15:42:47] <bodisiw> (okay! re-posting)
[15:42:54] <bodisiw> hi all, i have a text index on my field, but for some reason one term in a document doesn't match the document, while the others do... is there any special gotcha i should look for?
[15:43:00] <bodisiw> here's a gist demonstrating my issue https://gist.github.com/noahadler/e04bbacb4b4bab5ee5cb
[15:44:07] <marshall> hey mognodb
[15:45:08] <marshall> i just joined a project, and they're using this query: https://gist.github.com/anonymous/57909a40083c83695e8b. how do I add a clause that excludes all documents in the collection that have `type = coupon_redeemable`?
[15:47:54] <StephenLynx> http://docs.mongodb.org/manual/reference/operator/query/ne/
[15:51:34] <deathanchor> hmm.. anyone have an example of a commandline mongo comamnd for ensureIndex()?
[15:53:49] <Derick> db.col.ensureIndex ( { field: 1, field2: -1 } );
[15:53:52] <Derick> or rather:
[15:53:57] <Derick> db.col.createIndex ( { field: 1, field2: -1 } );
[15:54:45] <deathanchor> nah command line, like mongo --eval 'db.coll.ensureIndex({ whatever :1 })' database
[15:54:49] <deathanchor> would that work?
[15:55:11] <Derick> you need to add the db name
[15:55:16] <Derick> mongo dbname --eval ....
[15:55:19] <Derick> or
[15:55:22] <Derick> you did add the database name
[15:55:25] <Derick> at the end :)
[15:55:30] <deathanchor> yeah, let me test in my local instance
[15:55:31] <Derick> I think that should work
[15:57:28] <StephenLynx> gnome-terminal supports tabs
[15:57:39] <StephenLynx> plus I send some terminals that I don't need to interact to the secondary workspace
[16:02:38] <Cintax> probably stupid question, but how would I go about filtering out a nested document from a query? Because sometimes I want the nested doc, and sometimes I don't
[16:03:17] <StephenLynx> you just use projection of it like any other field.
[16:03:21] <StephenLynx> on it*
[16:03:40] <StephenLynx> of even on it's sub-values using dot notation. 'subdoc.value' : 1
[16:07:46] <Cintax> StephenLynx: ah ok, I was reading up on projection but I think I misunderstood it on my first read. Will give it a second look, thanks
[16:24:14] <drhalan> hey all. I am still looking how to store more complex spatial data ona flat 2d index
[16:34:07] <Bradipo> Anyone here use morphia with mongodb? It seems to be making a getMore query that I cannot account for. It is extremely slow, sometimes 10--300 seconds.
[16:35:25] <Bradipo> If I run the same query in the mongo shell, it's fast (like 100ms).
[16:35:44] <Bradipo> So there appears to be a difference in getMore and just a query.
[16:50:42] <deathanchor> Bradipo: that's reading more data from the cursor, in the shell the cursor only shows the first 10 docs I believe
[16:51:16] <Bradipo> Oh, does it?
[16:51:21] <Bradipo> I hadn't noticed...
[16:51:31] <Bradipo> Can reading from the cursor be avoided?
[16:51:38] <Bradipo> For example, like perhaps, using asList()?
[16:52:04] <Bradipo> Because apparently when it does this, it incurs huge delays.
[16:52:12] <Bradipo> When it uses the cursor that is.
[16:52:33] <Bradipo> Or is there likely something missing in the database that *should* just work?
[16:52:51] <Bradipo> To make the cursor read more performant.
[16:54:57] <Bradipo> What's odd, also, is that the code doesn't call any of the methods to return data...
[16:55:43] <Bradipo> It uses: for(Result r : query) {}
[16:56:00] <Bradipo> So where exactly does the query actually execute?
[17:55:49] <_ritchie_> anyone here use mongodb with django? I’m looking for help writing html forms for complex mongo documents (e.g. some fields are lists and dicts)
[20:19:17] <polydaic> I'm getting a readonly error while trying to set a Mixed type
[20:19:59] <polydaic> *sub property of a mixed type
[20:20:04] <polydaic> is there any way to prevent this
[21:40:41] <dman777_alter> if db.getProfilingLevel() shows level 2, shouldn't all queries be logged?
[22:58:53] <JuanDaugherty> there are some nice things in mongo
[22:59:49] <Epichero> my query is wrong, i'm using a framework but my problem is my query i think https://gist.github.com/Tyler-Anderson/798aa9e412237a453c35
[22:59:51] <Epichero> line 13
[23:00:11] <Epichero> i explain my problem about what i need to exclude