Evgeniy, Thanks for clarifying, I completely forgot about this behavior!
- Denis On Tue, Jun 16, 2020 at 5:10 PM Evgenii Zhuravlev <e.zhuravlev...@gmail.com> wrote: > Steve, > > Actually, disabling WAL is a good option for your use case. Checkpoint > mechanism is the same with disabled WAL, the only difference is that node > is not writing WAL to the disk on each operation. Usually, it might make > sense to disable WAL for initial loading - when you can lose the data in > case of failure and start data loading again. Four your use case, if you > don't care about restore, you can just disable it: > https://www.gridgain.com/docs/latest/developers-guide/persistence/native-persistence#disabling-wal > > Best Regards, > Evgenii > > > > вт, 16 июн. 2020 г. в 17:02, Denis Magda <dma...@apache.org>: > >> Steve, >> >> Please check these generic recommendations if you haven't done so >> already: >> https://apacheignite.readme.io/docs/durable-memory-tuning#native-persistence-related-tuning >> >> Otherwise, send us a note if you come across any bottlenecks or issues so >> that we can give you more specific recommendations. >> >> - >> Denis >> >> >> On Tue, Jun 16, 2020 at 3:25 PM steve.hostettler < >> steve.hostett...@gmail.com> wrote: >> >>> Thanks a lot for the recommendation. So keeping the WAL, disabling >>> archiving. >>> I understand all records are kept on disk. >>> >>> Thanks again. Anything else? >>> >>> >>> >>> -- >>> Sent from: http://apache-ignite-users.70518.x6.nabble.com/ >>> >>