Log file Viewer
#mongodb logs for Friday the 25th of September, 2015
[00:15:38] <darius93> Does mongo perform any locks? eg if i concurrently am writing (inserting and/or updating) a document within a collection, does it lock it so say thread b has to wait for thread a to finish its operation before it can write or update a document?
[01:16:10] <cheeser> but you could always try starting mongod with ssl turned on and see if it complains.
[05:59:25] <Jonno_FTW> in pymongo, If I use find(), iterate over the cursor, and change the document, how doI write that changed doc back to mongo? will it break the cursor?
[06:10:00] <Boomtime> Jonno_FTW: many drivers have a 'save' method and no, it won't break the cursor, though if you changed the values in the doc that your cursor is using to enumerate, there is the possibility of running over the same document again
[06:11:30] <Boomtime> be aware that the 'save' method, for those drivers which have such, is usually a full update - it replaces the entire existing document with the one you modified ('save' usually can't detect what you've changed)
[06:12:18] <Jonno_FTW> Boomtime: I've gone with update_one like this:collection.update_one({"_id": i["_id"]}, {"$set": {"anomaly_score": anomalyScore}})
[06:13:18] <Boomtime> you are only sending the data that is needed, and providing the server with the best possible chance for optimization
[06:13:31] <Jonno_FTW> just wondering because in other languages updating a collection while iterating it breaks the cursor like in java
[06:14:56] <Boomtime> mongodb isn't a language - the 'collection' on the server is concurrently accessible and you will see changes as they occur
[06:15:57] <Boomtime> if you run a query for a large number of documents, it is entirely possible that by the time you've enumerated towards the end, more documents have been inserted in the interim - and it is even possible your cursor will return some (all, or possible none) of the new documents that got inserted during that time
[06:16:40] <Boomtime> the only things you are guaranteed is that you'll never see a partial update to a single document
[07:11:39] <antiPoP> Hi, I'm trying to make a or search between text and number fields, but when I do it it says it's not indexed although there is an index there. What i'm doing wrong? Here are the commands and output? https://gist.github.com/antiPoP/a5e38acc07b631b257d0
[07:50:11] <AzaToth> So, I'm trying to make a small migration of a db, via some node-migration. the script in question can be found at https://paste.debian.net/313320/ and the db in question is a 2.4. The problem I have is that users.find() inside the forEach() doesn't give me any data at all on the server, but does so for me locally; thus I wonder if I can't call a second collection inside a forEach from a first one?
[07:57:47] <movedx> What could be the reason as to why 'mongod' won't start from the official init script provided by your RPMs, or a custom systemd service file I wrote, but it does start when I run it as root in a simple manner: mongod -f /etc/mongodb.conf
[07:59:01] <movedx> If I run it as root directly, it logs fine. If I run it using your init script, it doesn't log. It fails with absolutely no clue as to what is going on.
[07:59:20] <AzaToth> ax003d: though I notice the local execution isn't working either fully, it doesn't get the same number of users every run
[08:01:35] <AzaToth> ax003d: basically I want to set an array in "companies" containing all "users" who has a reference to said company in it's "owns" array, if said user's "needsInvite" flag is false
[08:02:26] <AzaToth> due to a added feature I want to update the database with initial logic based on that
[08:03:42] <AzaToth> even tried to run it from scratch as a plain script, without using node-migration, but getting the same issue, that not all "users" are found everytime
[08:04:22] <movedx> joannac: It's CentOS 7, so it rolls over to systemd, of course. I can't find the location that your RPM is putting the service file.
[08:11:23] <movedx> joannac: So obviously systemd supports init scripts, so systemd is executing the init script. Output is none existent. I've tried writing my own service file, but the same issue arises.
[08:12:01] <movedx> joannac: I'm going to try removing the user from the service file, to see if it starts as root from the service file.
[08:23:07] <movedx> I just don't get what I'm doing wrong here. I don't believe I've changed anything. Even from a fresh installed, minimal CentOS 7 box it doesn't work.
[08:25:02] <movedx> The executable exits with an exit code of '1': http://docs.mongodb.org/manual/reference/exit-codes/ -- wonderful.
[08:36:02] <movedx> And this is all I get out of it: ERROR: child process failed, exited with error number 1
[08:43:25] <movedx> "Failed global initialization: FileNotOpen Failed probe for "/data/db/mongodb.log": Permission denied" -- for the root user? Permission denied, for... the root user?
[08:44:13] <BurtyB> does the dir exist? are you using that horror called selinux? anything in the system logs?
[08:50:05] <movedx> Yes. Yes (why on Earth would you turn it off?). No. I'm going step-by-step through my Ansible code to determine if I'm doing something wrong, but really, all it does is download and install the RPM to the official repositories, installs MongoDB 3.0.5, drops the configuration onto disk (http://pastebin.com/h9GvxMPR), and starts the service which, of course,
[10:26:12] <jamiel> Hi all, I am trying to find a way to prevent remove actions on a secondary replica set member to act as an "archiving member" . Am I correct in thinking I can't set a user for the operations syncing the oplog to make use of privileges for this?
[11:15:27] <m3t4lukas> now I'm ready to do big data with mongoDB, I will never delete anything from now on :)
[18:22:36] <jmzc> db.Session.find({"events":{ $elemMatch: { "eventDescription":"CALL_REQUESTED", "eventTime" : { $gte : ISODate("2015-09-21T06:00:00Z") }, "eventTime" : { $lte : ISODate("2015-09-21T07:00:00Z") }}}});
[18:23:55] <jmzc> i'm only getting records before ISODate("2015-09-21T07:00:00Z") , and ignores "eventTime" : { $gte : ISODate("2015-09-21T06:00:00Z") }
[18:47:07] <jmzc> db.Session.find({"events":{ $elemMatch: { "eventDescription":"CALL_REQUESTED", $and : [ "eventTime" : { $gte : ISODate("2015-09-21T06:00:00Z") }, "eventTime" : { $lte : ISODate("2015-09-21T07:00:00Z") } ] }}});
[18:47:58] <daidoji> isn't it like "field1: {$and: [{ $gte : ISODate("2015-09-21T06:00:00Z") }, "eventTime" : { $lte : ISODate("2015-09-21T07:00:00Z") }]}" or something like that?
[18:49:30] <daidoji> db.Session.find({"events":{ $elemMatch: { "eventDescription":"CALL_REQUESTED", $and : [{ "eventTime" : { $gte : ISODate("2015-09-21T06:00:00Z")} }, {"eventTime" : { $lte : ISODate("2015-09-21T07:00:00Z") }} ] }}});
[19:04:34] <mbwe> i was wondering, what is the best way to store timeseries data with mongodb, or could somebody give me some pointers to soms docs dealing with this kind of data, or if mongodb is not the best choice, what other db i could use
[19:06:07] <daidoji> mbwe: I have not had a chance to use it, but I've heard good things about influxDB
[19:07:07] <daidoji> mbwe: no docs on hand, but if you think of each document as a "tick" in your time-series (assuming uniform distribution) then each document representing a tick with the _id set to the time it was taken seem like a good way to do things
[19:07:13] <mbwe> ok, but that is to say that mongodb is not a good fit for timeseries daidoji ? i prefer mongo to be honest
[19:07:49] <daidoji> if you want to be able to do adhoc analysis and querying over your time series, its probably not the best fit imo
[19:08:21] <mbwe> http://blog.mongodb.org/post/65517193370/schema-design-for-time-series-data-in-mongodb
[20:26:06] <[diecast]> having some issues. both of my secondaries in a replication set are down and have oplog too far out of sync errors. i stopped them and started to work on the primary when the primary now says it is a secondary
[23:10:18] <macwinner> hi, any ideas why my query stream for mongo prints out only the first record of my returned data here: var stream = Activity.find(filter).read("primaryPreferred").stream(); stream.('data', console.log);