[jira] [Comment Edited] (KUDU-3452) Support creating three-replicas table or partition when only 2 tservers healthy

["Alexey Serbin (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/KUDU-3452?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17702870#comment-17702870 ] 

Alexey Serbin edited comment on KUDU-3452 at 3/20/23 6:15 PM:
--------------------------------------------------------------

Having a Kudu cluster with just 3 tablet servers is already quite limiting -- if any of these 3 nodes fail, there isn't a place for replicas which are to migrate from the failed node.  So, a minimum recommended number of tablet servers nodes is a Kudu cluster is 4 already.  With RF=3, that allows to keep Kudu data highly available while operators can bring the failed node back.

Allowing to create a table with RF=3 when only two tablet servers are alive is asking for a disaster, because the table's data will become non-available when just one more node goes down.

Why to create tables with RF=3 then if just a single node failure renders your data unavailable?  Since it doesn't provide a benefit of high availability, you could create tables with RF=1 and have just one tablet server in your cluster.


was (Author: aserbin):
Having a Kudu cluster with just 3 tablet servers is already quite limiting -- if any of these 3 nodes fail, there isn't a place to replicas which are to migrate from the failed node.  So, a minimum recommended number of tablet servers nodes is a Kudu cluster is 4 already.

Allowing to create a table with RF=3 when only two tablet servers are alive is asking for a disaster, because the table's data will become non-available when just one more node goes down.

Why to create tables with RF=3 then if just a single node failure renders your data unavailable?  You could create tables with RF=1 and have just one tablet server in your cluster.

> Support creating three-replicas table or partition when only 2 tservers healthy
> -------------------------------------------------------------------------------
>
>                 Key: KUDU-3452
>                 URL: https://issues.apache.org/jira/browse/KUDU-3452
>             Project: Kudu
>          Issue Type: Improvement
>            Reporter: Xixu Wang
>            Priority: Major
>
> h1. Background
> In my case, every day a new Kudu table (called: history_data_table) will be created to store history data and a new partition for another table (called: business_data_table) to be ready to store today's data. These tables and partitions all require 3 replicas. This business logic was implemented by some Python scripts. My Kudu cluster contains 3 masters and 3 tservers. Flag: --catalog_manager_check_ts_count_for_create_table is false.
> Sometimes, one tserver maybe become unavailable. Table creating task will retry continuously and always fail until the tserver become healthy again. See the error:
> {color:#ff8b00}E0222 11:10:32.767140 3321 catalog_manager.cc:672] Error processing pending assignments: Invalid argument: error selecting replicas for tablet 41dffa9783f14f36a5b6c35e89075c1a, state:0: Not enough tablet servers are online for table 'test_table'. Need at least 3 replicas, but only 2 tablet servers are available{color}
> {color:#172b4d}As there are no enough replicas, a tablet will never be created. The state of this tablet is not running. Therefore, read or write this tablet will fail even if there are 2 tservers can be used to create 2 replicas.{color}
>  
> An already created tablet can still be on service even if one of its 3 replicas become unavailable. Why can not create a three-replicas table when only 2 tservers healthy?
>  
> h1. Design
> A new flag: --support_create_tablet_without_enough_healthy_tservers is added. The original logic keeps the same. When this flag is set true, a three-replicas tablet can be created successfully and its status is losing one replica. This tablet can be be read and write normally.
>  
> There are 3 things need to do:
>  # A tool to cancel the table creating task.
>  # A tool to show the running table creating task.
>  # A method to create table without enough healthy tservers



--
This message was sent by Atlassian Jira
(v8.20.10#820010)