Skip to content

Bug Report: PRS fails with "Data too long" error #13286

@deepthi

Description

@deepthi

Overview of the Issue

When attempting to run PlannedReparentShard on a cluster with long-ish tablet aliases, it throws the following error:

failed: Data too long for column 'primary_alias' at row 1 

From the logs, it's clear that the error is coming from

INSERT INTO _vt.reparent_journal (time_created_ns, action_name, primary_alias, replication_position) VALUES <redacted>

The current column definition is varbinary(32). It should be increased to varbinary(250) or even 255.

Reproduction Steps

Create a test vitess cluster with a tablet alias > 32 characters.
Attempt to PRS to that tablet.

Presumably this will also fail ERS (EmergencyReparentShard), but I haven't tried it.

Binary Version

Initially found with v15.0.3, but I've verified that every version is affected up to current `main`.

Operating System and Environment details

N/A

Log Fragments

No response

Metadata

Metadata

Assignees

Type

No type
No fields configured for issues without a type.

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions