Log file Viewer

[18:00:19] <c4017w> which aggragation can I use to split a document with multiple keys into multiple documents, one for each key?

[20:55:27] <BeryCZ> Hello, can anybody help me with this? I am desperate

[20:55:27] <BeryCZ> https://developer.mongodb.com/community/forums/t/mongod-config-wtcheck-tthread-writes-on-disk-into-wiredtiger-wt-every-minute/14907

[20:55:27] <BeryCZ> basically mongodb writes 40kB on disk every minute even tho I do no changes. It saves the WiredTiger.wt, WiredTiger.turtle and an index file

[20:55:27] <BeryCZ> Can someone exmplain why does it do that or tell me how can I prevent it?

[21:12:18] <aimorris> I'm trying to write a trigger function in my Atlas application to run every week, but I can't get the "service" as it returns undefined... I'm trying to do it via

[21:12:29] <aimorris> const collection = context.services.get('mongodb-atlas');

[21:12:46] <aimorris> For the serviceName, I've also tried Cluster0 and mongodb-datalake

[21:56:56] <d4rkp1r4t3> c4017w: $unwind maybe?

[22:00:19] <d4rkp1r4t3> BeryCZ: WT journal does a checkpoint every 60 seconds, so its probably that. Its probably updating some kind of metadata about last commit and checkpoint. You can disable jounaling on standalone instances, disable it with caution. https://docs.mongodb.com/manual/core/wiredtiger/#journal

[22:06:45] <c4017w> d4rkp1r4t3, that's what I ended up doing, but I don't like it. Need to $objectToArray first, and after $unwind, it's no longer an array, so I can't easily $arrayToObject... Whatever, it's good enough

[22:23:16] <d4rkp1r4t3> not to mention with large arrays its going to be inefficient. might consider schema change

[22:51:04] <BeryCZ> d4rkp1r4t3: storage.journal.enabled:false, it is off, but still doing that, I dont really see a reason for any writes since there is no client even connected to the mongodb server :-/

[23:36:00] <c4017w> Anyone familiar with the node-mongodb-native driver? I'm using collection.aggregate(pipeline, {raw: True}) to return raw BSON. This seems to return data in small (~6MB) batches regardless of any batchSize settings. Any way to disable this?

[23:36:58] <c4017w> I'm piping the cursor directly to an HTTP response, but the stream always emits 'end' after the first chuck

[23:41:26] <c4017w> Where can I find documentation on this anyway? The only reference I found to the 'firstBatch' thing returned by AggregationCursor was on some stackOverflow post...