Log file Viewer
#mongodb logs for Thursday the 17th of October, 2013
[00:10:36] <zmansiv> if i have "project" documents and each project can have multiple screenshots, would it be better to a) store them as an array of binary data in the project document, b) store them as an array of base64d strings in the project document, or c) use gridfs?
[02:30:14] <geoffeg> If I have a huge DB, say 500GB, and I run a db.coll.remove({'some-non-indexes-field' : 1}) in mongodb 2.2, will that grab the db write lock and prevent other writes?
[02:43:21] <TkTech> geoffeg: For our huge (multi terabyte) collections, we gave up using MongoDB for anything other than raw storage.
[02:43:45] <TkTech> geoffeg: We crawl products, including their attributes and other metadata and write them into a sharded MongoDB database
[02:44:16] <TkTech> geoffeg: Then periodically read (from the replicas only) into redshift and do our work there.
[05:06:22] <Steve009> i get: Database command 'aggregate' failed: (errmsg: 'exception: invalid operator: $eq'; code: '10068'; ok: '0.0'). (Mongo::OperationFailure) with the $eq
[05:38:44] <joannac> I would be a bit dubious that such an bug would've gone unnoticed though. Steve009 -- what version of the ruby driver?
[05:41:51] <Steve009> will have to test it with robomongo later to see if it works with straight mongo commands
[06:02:58] <deepender> now i have to add another user also who will alos have their different folder
[08:25:59] <Tiller> Is it possible from {id: 1, a: "Last", b: ["Old", "Old"]} to make an update request to have {id: 1, a: "Last", b: ["Old", "Old", "Last"]} ? =/
[09:00:00] <Tiller> Thanks joannac, I saw something like that before :) I'll have to look at the java driver doc :)
[09:46:09] <DragonBe> what's the preferred setup for running a single mongodb instance? disk, RAM and CPU requirements are our major focus at this time.
[12:07:55] <fel> haha!! and the first M101J (starting on 21st Oct.) as the second M101JS (starting on 14th Oct) have the dead python picture instead of coffee and nodejs logos
[12:16:37] <joannac> Okay, I can't fix it, but I pinged some people. WIll probably be fixed once the US wakes up.
[12:20:35] <tonni> how do I query for the last element of an array? Something lihe the first element from the docs: db.inventory.find( { 'memos.0.by': 'shipping' } )
[14:41:58] <quickdry21> I'm having some weird issues with primaries refusing/closing connections... When I do a rs.status() from a secondary, the primary is shown as healthy, with the last heartbeat received w/in the last minute... I'm getting the following error when trying to connect via mongo shell:
[14:41:59] <quickdry21> Error: DBClientBase::findN: transport error: giordano.gmlapi.com:10002 ns: admin.$cmd query: { whatsmyuri: 1 } at src/mongo/shell/mongo.js:147
[15:56:28] <stongo1> I have three servers, each with host names in /etc/hosts pointing to each other. I'm trying to create a replica set with the three machines, but keep getting an error on rs.initiate saying all members are not up
[15:58:48] <joe_p> stongo1: they all have to be able to reach each other. try doing a telnet to the mongo port from each machine to make sure they are all reachable from one another
[16:07:48] <TkTech> saml: By now, likely not. Why were you using it in production with no redundancy or backups? :|
[16:12:28] <TkTech> There are a few methods to recover a file (assuming it hasn't been 0'd) depending on the filesystem and OS.
[16:13:31] <stongo1> hrm, looks like I had to turn auth off on the mongo shell I was trying to execute rs.initiate() from
[16:14:14] <saml> i started new mongod using the test.0, test.ns, journal/ but test collection in the test db is gone :(
[16:15:21] <stongo1> my admin user has clusterAdmin privs ... that's weird rs.initiate() wouldn't work
[16:32:23] <stongo1> ugh, replica sets were much easier to setup in the mongodb course when each instance was on the same machine :P
[16:35:30] <stongo1> seems like auth is creating the problem ... disabling auth allows me to add the other machines
[16:36:30] <stongo1> so how does one use auth with replica sets? should the rs.add() string be "username:password@name:port/dbname" ?
[17:12:10] <stongo1> I added a keyFile to mongodb.conf, but now don't have any permissions to do anything when I re-enter mongo shell
[17:13:02] <stongo1> the example of setting up a replica set on mongodb doesn't use any auth, so pretty much useless in the real world
[17:33:02] <saml> Nodex, give me a script that parses mycollection.* files and gives me json documents so that i can import them again
[19:01:03] <redsand> is there a format that mongoimport or mongorestore needs for writing bson objects directly to a file, and then importing them in?
[21:03:17] <duncancmt> Hi! After a reboot, I'm getting this error message on my sharded cluster, and I can't access any of my data. "error creating initial database config information :: caused by :: can't find a shard to put new db on" the code is 10185.
[21:04:56] <duncancmt> why?! How the hell is that acceptable behavior for a database?!?! It wasn't even an unclean shutdown
[21:24:31] <joe_p> duncancmt: are your config servers running - the error log implies they are not - ERROR: config servers not in sync! no config servers reachable
[21:27:16] <duncancmt> yeah, the config servers took a bit longer to come up than I expected and I jumped the gun on starting the mongos