Log file Viewer

[12:31:16] <dusted> I have a sharded cluster, and I'm wondering.. Would the unordered bulk operation (with .find(id).deleteOne() ) be preferred over a single collection.deleteMany( { id: { $in: idList } } ); or are they totally the same? the deleteMany seems to run faster "on my machine"

[12:31:26] <dusted> I can't test it on the cluster atm

[12:32:11] <dusted> something tells me that the bulk operation ought to be able to split calls and send to shards and secondaries that need it, while the deleteMany with the list would have to send the entire list to everyone, making it less efficient, but I'm not sure if that's true

[13:47:07] <guardian> hello, is it a good or bad idea to use transactions and abort_transaction() to implement a dry run feature for a multi collection update?

[18:23:17] <lifelearner> When is this channel usually active?

[18:27:07] <dusted> ‰Mostly never

[19:37:36] <ekaj> ~dustyou scared him off

[19:37:40] <ekaj> * dusted

[20:50:30] <kenperkins> so, am I crazy for trying to use nfs for mongo?

[22:47:01] <kenperkins> well, since nobody answered, I decided it _was_ crazy, so instead i figured out how to iscsi working for it

[23:27:56] <dusted> kenperkins: yes, I'm not sure if there's any sane use for nfs ;) :P

[23:28:07] <dusted> mongo recommends xfs

[23:28:50] <dusted> that could be on iscsi though, i guess :)

[23:29:31] <kenperkins> xfs is filesystem?

[23:29:37] <dusted> yes

[23:30:13] <kenperkins> currently it's ext4