Currently, we are developing SPILL DISK now, you can refer the code from this branch. https://github.com/apache/doris/tree/spill_and_reserve This feature is almostly finished, we are doing some tests now. You can use it in next release.
Joice Jacob <joicejacob1...@gmail.com> 于2024年12月10日周二 12:37写道： > > Dear Apache Doris Development Community, > > Our team is very satisfied with the performance of Apache Doris, but we > have a few concerns that we would like to clarify. > > We are planning to replace Hive with Doris for our use cases. In Hive, > queries typically complete successfully using the available resources, even > under constrained conditions. Most of our queries are small; however, we > have some offline reporting use cases that require computing data for 3, 6, > 9, or 15 months. > > In Doris, we have observed that queries fail when resources are > insufficient. This poses a challenge, as scaling the cluster to accommodate > large queries isn't feasible due to their infrequent execution. Even when > the Doris cluster is entirely free, large queries sometimes fail due to > resource limitations. > > Our question is: > > > 1. Is there any configuration in Doris that can ensure queries complete > successfully when resources are constrained, rather than failing? > 2. Can we expect a feature in future releases that allows large queries > to gracefully utilize available resources and complete without failing, > even when memory is limited? > > We appreciate your insights and guidance on this matter. > > Best regards, -- Best Regards, Yi Guo Lei SelectDB The Best Cloud Native Platform for Apache Doris Tel:134-3991-0228 Email:yiguo...@gmail.com --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@doris.apache.org For additional commands, e-mail: dev-h...@doris.apache.org
