Log file Viewer

[10:28:21] <AppStore> Master slave replication is deprecated in favor of replica sets according to the documentation, but I have a use case that seems to fit master slave better than replica sets: I have one master application which gathers data, then needs to distribute it to api nodes which in turn serves it read-only to android/ios applications. Is there a strong reason to

[10:28:22] <AppStore> still avoid old style master/slave replication?

[10:34:16] <Derick> AppStore: you can do the same with ReplSets too - just mark all the secondaries as hidden nodes

[10:41:37] <daumie> How did you guys become pros in Mongo

[12:20:55] <JustinHitla> what is #velociraptor about ? there is no one in here

[12:24:24] <StephenLynx> I guess is just something about the bot being used to log the channel.

[12:37:14] <shlant> hi all. I have a weird issue with mongoose and mongo 3.2.4. when I start my app, I see "Successfully authenticated as principal user1 on admin"

[12:37:26] <SteamWells> shit, friday and i just created a pants index on a huge collection, and its my last day *RUNS*

[12:37:34] <StephenLynx> Don't use mongoose, shlant

[12:37:39] <StephenLynx> its nothing but trouble.

[12:37:47] <shlant> but then when my app tries to run an aggregate command, I get "SCRAM-SHA-1 authentication failed for csm-policy-watcher on admin from client 10.0.0.76 ; UserNotFound: Could not find user user1@admin"

[12:38:14] <shlant> @StephenLynx yea, it seems the driver for node in general are junk, especially with TLS

[12:38:27] <StephenLynx> the standard driver is alright.

[12:38:34] <shlant> but why would it be able to auth user1@admin on startup, but not on aggregate?

[12:39:32] <StephenLynx> maybe the user isn't allowed to query the collection you are trying to.

[12:39:47] <shlant> it has dbOwner perms

[12:40:25] <shlant> let me try the command amnually

[12:45:25] <shlant> it seems to run manually

[12:45:35] <shlant> maybe that's not what's throwing the missing user

[12:48:56] <StephenLynx> maybe you are trying to query using a new connection that haven't been authenticated?

[20:39:37] <jamieshepherd> Hey. Got a 1.7gb mongod.log, which seems to just be logging writes and updates (no errors). wondering if this is misconfigured somewhere

[20:39:43] <jamieshepherd> and how I might configure it not to do that :D

[21:55:34] <cjhowe> sup guys, trying to understand the rationale behind this part of the spec: https://github.com/mongodb/specifications/blob/master/source/server-discovery-and-monitoring/server-discovery-and-monitoring.rst#network-error-when-reading-or-writing

[21:55:57] <cjhowe> why can't writes ever be retried? if the primary crashed and closed the connection, shouldn't it wait for the next primary election for a while before giving up?

[21:56:14] <cjhowe> is that just because writes are more likely to need to handle the error in other ways?

[22:10:22] <cheeser> primarily becuase retry semantics are tricky to do generically and correctly

[22:11:25] <cjhowe> yeah, that makes sense, thanks

[22:12:47] <silviolucenajuni> Hi

[22:13:13] <silviolucenajuni> someone have troubles with a query that use equality and $in and need a $sort.

[22:13:43] <silviolucenajuni> I have a collection with 10 billions of documents, and cannot create a Index to cover a query that use $sort, $in and equality.

[22:13:50] <silviolucenajuni> ???

[22:18:25] <silviolucenajuni> someone have troubles with a query that use equality, $in and need a $sort in a very big collection?