MR3

1. Could you share the logs of HiveServer2 and Metastore when you execute 'show databases'?

It has been replaced with MR3 Cloud, which is a web interface for running Hive on public clouds as

Alright, we will try to update JDBC storage handler in Hive-MR3. You are welcome to submit PRs to the

Could you try Hive-MR3 with autoscaling enabled? With autoscaling, MR3 assume infinite resources on

great. On Sunday, 3 April 2022 at 16:40:52 UTC+8 Sungwoo Park wrote: A new blog article is available:

We are planning to use SPARK ON MR3, Because we want to use the features of mr3 to optimize resource

I have a small question. Why can't the FAQs page be found on the homepage of datamonad? Do you

I found that when there are a large number of small files in HDFS, the initialization speed of mr3 is

That sounds great! I'm looking forward to it. On Thursday, 17 February 2022 at 15:01:39 UTC+8

Through experiments, I have determined that the reason why am cannot allocate resources is : Can mr3

Thanks! On Tuesday, 18 January 2022 at 23:52:48 UTC+8 Sungwoo Park wrote: I uploaded a pre-release of

Thank you! On Friday, 14 January 2022 at 18:04:18 UTC+8 Sungwoo Park wrote: Attached is mr3.json

Hello, Hive-MR3 reads tez-site.xml, but it does not create tez-conf.pb because all configuration

Currently there is a problem with upgrading Log4j 2 to 2.15.0, so you could use a quickfix by adding

Limits.setConfiguration() is called in MR3 DAGAppMaster (and MR3 ContainerWorker), so it is not the

Actually there is no official release of Hive 3.1.3, and there is no additional Hive dependencies.

We have found that Hive 3 on MR3 1.3 shows abnormal behavior caused by two Hive patches (HIVE-20703

Hi Sungwoo, Thank you. That was the issue. MR3 doc : https://mr3docs.datamonad.com/docs/k8s/advanced/

We are pleased to announce the release of MR3 1.3. Highlights in this release are: 1) MR3 On both

MR3 1.3 will be released soon, and for those who want to try Hive 3 on MR3 1.3 on Kubernetes, we have

This issue has been fixed in MR3 1.3. It turns out that the fix shown above is a complete solution (

In the current implementation of MR3, you cannot specify node affinity rules for master and worker

You can use the new Docker image (mr3project/hive3:1.3) without changing anything in your scripts. Or

By the way, my email is cdmikechen@hotmail.com . If you can provide the working version, you can send

In MR3 1.3, we will introduce `mr3.auto.scale.out.num.initial.containers` which specifies the number

Actually, from MR3 1.3 (to be released next month), there will be no limit in the aggregate memory of

Thank you, Sungwoo. The issue was with the security group for LB. Added port 443 to the allowed list

This is probably because the label node-role.kubernetes.io/master is attached to the node. From https

MR3 does its best to avoid OOM in PipelinedSorter and reduce delays in (by trying to reuse existing

Hi, This error says that another process is using the port for Web UI. 2020-12-02T10:53:00336 ERROR [