Log file Viewer

[00:58:10] <flaf> Hi @all, I have a replica set with 3 hosts (mongo 2.4). If 2 servers are down, the mongo cluster is down. If only one server is down, the cluster is up. Is it normal? Is there a concept of quorum with mongo? (I'm a mongo beginner).

[00:58:45] <cheeser> you've just captured it really.

[00:59:05] <joannac> https://docs.mongodb.org/manual/core/replica-set-elections/

[00:59:10] <cheeser> as long as the majority of members is up, you can write

[00:59:38] <joannac> and possibly https://docs.mongodb.org/manual/core/replication-introduction/

[01:01:02] <flaf> Ah ok, ok. thx guys.

[01:01:40] <flaf> But readings is still possible with only 1 server up (among 3), correct?

[01:01:56] <cheeser> with seconadary reads, yes

[01:25:10] <flaf> thx for your helps. ;)

[02:00:43] <flaf> I have noticed that mongodb (2.4) is very verbose in my logs. I have a web site connected to my replica set and each time a web page is loaded I have lines in the logs (“connection accepted”, “end connection”. However I have put “setParameter = logLevel=0” in mongo.conf.

[02:01:04] <flaf> Is it possible to avoid these logs at each connection?

[02:01:36] <flaf> Or the next step for me is to just stop logs?

[02:18:14] <flaf> I have tried too “setParameter = logLevel=0” _and_ “setParameter = quiet=1”, but I still have logs at each connection.

[02:48:56] <joannac> flaf: you shouldn't after setting quiet:1

[02:50:30] <flaf> ah? Is the syntax “setParameter = quiet=1” correct in mongo.conf?

[02:50:40] <flaf> I have a doubt.

[02:55:42] <joannac> flaf: did you restart after making that change?

[02:56:00] <flaf> of course. ;)

[02:56:24] <flaf> should I set this on each node?

[02:56:48] <joannac> depends if you want it in effect on each node

[02:57:09] <flaf> ok.

[02:58:08] <flaf> I have a doubt concerning the syntax.

[02:58:39] <flaf> Is it “setParameter = quiet=1” or “setParameter = quiet:1”? (in moobot.conf)

[02:59:44] <joannac> should be setParameter = quiet=true

[02:59:49] <joannac> https://docs.mongodb.org/v2.4/reference/configuration-options/#quiet

[07:10:51] <flaf> joannac: sorry for my latency. So yes indeed with “setParameter = quiet=true”, it works. Thx very much for your help. ;)

[07:49:44] <flaf> however, I still have lines like "authenticate db".

[08:41:00] <flaf> je vois dans les logs une tonne de connexions de sqy (dans mongo)

[08:41:21] <flaf> au passage, j'ai du mal à le faire taire ce connard.

[08:43:30] <flaf> (ooops bad chan, sorry)

[09:00:56] <flaf> If you know how to disable “[connXXX] authenticate db: foo” in the log with mongodb 2.4, I'm really interested. Because I have set “setParameter = quiet=true” and “setParameter = logLevel=0” and the line are still here.

[09:19:16] <chrisq> is there a shortcut to run a query found in currentOp()?

[13:55:42] <m3t4lukas> seems like monday is everywhere :D

[13:56:34] <cheeser> sounds like a song by the Mammas and the Pappas

[14:01:16] <m3t4lukas> cheeser: #unintended

[14:03:33] <m3t4lukas> http://youtu.be/_ayy8IMPCUU

[14:13:52] <m3t4lukas> cheeser: is there a way to project a random number into an aggregation?

[14:14:17] <Zelest> o/ m3t4lukas

[14:14:30] <m3t4lukas> o/ Zelest

[14:14:39] <Zelest> finally got my 3 nodes up :D

[14:14:41] <Zelest> firewalled and sorted

[14:14:58] <m3t4lukas> Zelest: yaaaay, that's great news ;) I'm happy for you

[14:15:05] <Zelest> thanks :D

[14:15:13] <Zelest> now I need to sort out the replicaset part

[14:15:15] <Zelest> and make them love eachother

[14:15:51] <m3t4lukas> yeah, show them some of that good old mongodb love :D

[14:16:05] <Zelest> :)

[14:18:25] <cheeser> m3t4lukas: i don't know of one offhand

[14:19:08] <m3t4lukas> cheeser: it would be kinda important :/

[14:19:23] <cheeser> you could always file a jira for that

[14:19:32] <cheeser> they're not hard to write apparently

[14:19:37] <m3t4lukas> cheeser: or I could do it myself :P

[14:19:47] <cheeser> patches are even better :)

[14:20:12] <cheeser> our c# guy learned how and wrote 2 over the course of a few days

[14:40:37] <m3t4lukas> cheeser: I really need that feature so I'll do it now. Is github your main source tree or do you have public inhouse gitlab?

[14:42:34] <m3t4lukas> cheeser: am I assuming right that it is placed most appropriate in the projection pipe?

[14:43:23] <cheeser> github

[14:43:53] <m3t4lukas> cheeser: I'd like to coordinate it a bit so it get's into the releases as quick as possible so I don't need to set up an infrastructure for deployment of my own branch of mongodb

[14:46:23] <cheeser> file a jira so the kernel team knows about. add a note that you're working on a patch.

[14:47:57] <cheeser> https://github.com/mongodb/mongo/blob/master/CONTRIBUTING.rst

[14:48:22] <m3t4lukas> thanks a lot cheeser

[14:48:26] <cheeser> np

[14:56:29] <m3t4lukas> cheeser: how can I assign it to me?

[14:56:41] <m3t4lukas> *it = the issue

[15:00:10] <cheeser> you can't

[15:05:36] <m3t4lukas> cheeser: okay, doesn't stop me from working on it. I found out what I need to do it. Now I need to run some tests. Do you have a build environment on a vm?

[15:18:39] <cheeser> i don't really build the db

[15:24:18] <m3t4lukas> cheeser: okay

[16:03:56] <saml> i created mongodb. all thank me.

[18:42:57] <morf> mongodb is really jolly good fun

[18:43:51] <morf> i have work done like 5 times faster than with traditional dbs

[18:44:40] <cheeser> that's what we like to hear :)

[19:52:09] <cornfeedhobo> does anyone know what port mongo initiates replication/clustering on? have a node which has opened up 27017:27019 but still i am required to drop my firewall to begin peering

[19:53:06] <Derick> it uses 27017

[19:53:27] <Derick> for replication at least

[19:54:02] <cornfeedhobo> yeah, once they know about each other, that is the case, but it appears that the initiating is not done on it

[21:19:46] <ThisIsDog> @joannac: I looked into the $and issue you helped me with before the weekend. We think we think the problem is with Ming's implementation of mongo in memor, and Ming has been very unresponsive with us in the past. How stable is Mongo's in memory storage engine? Should it be stable enough for us to use running tests?

[21:57:38] <Ryan___> Hello.

[22:00:19] <m3t4lukas> hey Ryan___

[22:06:19] <magicantler> Can you get a speed improvement by sharding on the same server?

[22:08:13] <cheeser> probably not. each instance would be competing for resources

[22:08:41] <cheeser> though depending on the size of your dataset you might get some parallelization benefits.

[22:11:43] <magicantler> Cheeser: What if it's 3 massive machines, with 128gb ram each?

[22:12:35] <magicantler> cheeser: this SO post, says clearly not, but it's a few years old http://stackoverflow.com/questions/9203418/mongodb-sharding-on-one-machine

[22:14:28] <cheeser> i'd be surprised, tbh

[22:16:11] <magicantler> cheeser I would guess that shards on a single machine still parallelizes write and reads

[22:16:22] <magicantler> is that true?

[22:18:10] <cheeser> yes, they would but they'd be fighting the other node(s) for memory and cpu time

[22:18:28] <cheeser> so you might gain in one area only to lose in another. how that balances out is hard to predict

[22:20:56] <magicantler> Thanks.. I guess I'll have to experiment with predicted load and data capacity. Is it tough to move a shard over to a new machine from a pre-existing one?

[22:23:16] <cheeser> i've never tried but i'd think it'd be the same procedure as moving a replica set node: bring up a new node, let it sync fully, decommission the old node.

[22:26:41] <L00tefisk> any idea why it wont let me insert into the database here? http://hastebin.com/eqagagizuh.py

[22:27:06] <cheeser> depends on the error

[22:28:11] <L00tefisk> i was gonna post it but it disappeared

[22:28:14] <L00tefisk> json.decoder.JSONDecodeError: Expecting value: line 1 column 1 (char 0)

[22:28:16] <L00tefisk> there we go

[22:29:13] <L00tefisk> if i print it to console it has a b infront of the request.data

[22:29:44] <L00tefisk> s has, that is