PMXBOT Log file Viewer

Help | Karma | Search:

#mongodb logs for Tuesday the 10th of December, 2019

(Back to #mongodb overview) (Back to channel listing) (Animate logs)
[00:36:45] <brandon-> Hello, I am having an issue w/ results returned via a pyMongo find() with $elemMatch projection. roughly 80% of the records returned are the document _id field and the matched sub-domain in the array. HOWEVER, the remaining 20% are *JUST* the document's _id and that's it. i checked the document via mongo shell and indeed it has the matched array. the query is like so:
[00:37:05] <brandon-> ams = db.filedb.find( {'file.owner': '0', }, { 'file': { '$elemMatch': { 'deleted': False } }, }, )
[00:38:20] <brandon-> ^- as i said ~80% of the returned records are the document _id and the 'file' array element match - the full sub-document matched with 'deleted': False and 'owner': 0. but the other 20% are lacking this and have the _id of the document only. i am confused as to how this is happening?
[23:06:19] <blizzow> I've been doing mongodumps successfully for quite some time and then today they started failing with this...
[23:06:32] <blizzow> Failed: error writing data for collection `mydb.mycollection` to disk: error reading collection: (CursorNotFound) cursor id 291880998508 not found
[23:14:28] <GothAlice> Sounds like a possible server-side cursor timeout. No idea why that might have changed other than data growth beyond some threshold for time it takes to back up.
[23:14:42] <GothAlice> blizzow: ^
[23:17:56] <blizzow> GothAlice, I think I figured it out. I was using the latest mongodump (4.2) against a 3.6 database.
[23:18:04] <blizzow> Weird.
[23:24:41] <blizzow> I would think mongodump would be backwards compatible.