[CASSANDRA-16411] Fix topology corruption on joining nodes without DC in dtests - ASF JIRA

Log work

Agile Board

Rank to Top

Rank to Bottom

Attach files

Attach Screenshot

Bulk Copy Attachments

Bulk Move Attachments

Voters

Watch issue

Watchers

Create sub-task

Convert to sub-task

Move

Link

Clone

Labels

Update Comment Author

Replace String in Comment

Update Comment Visibility

Delete Comments

XML

Word

Printable

JSON

Details

Type: Bug
Status: Resolved
Priority: Normal
Resolution: Fixed
Fix Version/s: 4.0-rc1, 4.0
Component/s: Test/dtest/python
Labels:
None

Bug Category:
Correctness
Severity:
Normal
Complexity:
Normal
Discovered By:
User Report
Platform:

All
Impacts:

None
Since Version:

NA
Source Control Link:

8762267b13fffb921ae79b0220630f10695ee9d9 https://github.com/riptano/ccm/commit/e5870fd081a927b6d4bad47719b1b1f893d70c3e https://github.com/apache/cassandra-dtest/commit/ad9462ee737ebc92a86cb5b266fccdd93e1111a9
Test and Documentation Plan:

Hide

See ticket

Show
See ticket

Description

Dtests with multi-dc setups don't update the topology properties file of a newly joining node without dc post cluster creation.

The new node will join dc1 but it's snitch will report the default snitch config as it doesn't get updated == dc1 == all nodes are in dc1 == there are no other dcs. That leads to bootstrap, token, startup,... problems. Let's see an example:

populate [1, 1, 1]
Node1 sees: dc1/ip1, dc2/ip2 and dc3/ip3
Node2 sees: dc1/ip1, dc2/ip2 and dc3/ip3
Node3 sees: dc1/ip1, dc2/ip2 and dc3/ip3
Add a new Node4
Node4 sees: dc1/ip1+ip2+ip3+ip4

All multi-dc dtests that add a node after cluster creation suffer this problem.

Solution: When a new node is added without specifying a dc in dtests check for existing dcs. If there are any force to specify one to update the topology file.