[MCOL-3729] ExeMgr and PrimProc reconnect Created: 2020-01-16  Updated: 2023-10-25  Resolved: 2023-10-25

Status: Closed
Project: MariaDB ColumnStore
Component/s: ExeMgr, PrimProc
Affects Version/s: 1.5.3
Fix Version/s: Icebox

Type: Bug Priority: Major
Reporter: David Hall (Inactive) Assignee: Unassigned
Resolution: Fixed Votes: 1
Labels: None

Issue Links:
Relates
relates to MCOL-4015 ExeMgr must re-establish its PrimProc... Closed

 Description   

Currently, ExeMgr and PrimProc don't attempt to reconnect after comm failure. Specifically after one or the other Segv. ProcMon currently sees the failure of one or the other and restarts both. This is clumsy. With the move of control to other mechanisms, asking them to do this is unreasonable. In addition, if the com just happens to fail temporarily, there may be no attempt to restart, and certainly no attempt to reconnect

This Jira is to create awareness of the comm state and automatic reconnect between the two processes.

A check of other procerss, such as DMProc and DDLProc needs to also be done. The problem may be more widespread.

The solution may be to create a class for all processes to use to help re-establish connections.



 Comments   
Comment by Doug Whitfield [ 2023-08-02 ]

@roman should this be closed since https://jira.mariadb.org/browse/MCOL-4015 was closed, or is this a separate issue?

Generated at Thu Feb 08 02:45:00 UTC 2024 using Jira 8.20.16#820016-sha1:9d11dbea5f4be3d4cc21f03a88dd11d8c8687422.