Repository navigation
[Neo4j] Default wait strategy breaks on IPv6-enabled environments #4454
Description
The JDBC driver project I am working on recently got build errors with TestContainers (with Neo4jContainer) tests in our TeamCity instance, that looked like this:
org.testcontainers.containers.ContainerLaunchException: Container startup failed
org.testcontainers.containers.ContainerLaunchException: Container startup failed
Caused by: org.rnorth.ducttape.RetryCountExceededException: Retry limit hit with exception
Caused by: org.testcontainers.containers.ContainerLaunchException: Could not create/start container
Caused by: org.testcontainers.containers.ContainerLaunchException:
Timed out waiting for log output matching '.*Bolt enabled on 0\.0\.0\.0:7687\.After dumping the logs via withLogConsumer(outputFrame -> System.err.print(outputFrame.getUtf8String())), something interesting appeared:
2021-09-08 15:38:05.635+0000 INFO Starting...
17:38:09 2021-09-08 15:38:09.752+0000 INFO ======== Neo4j 4.3.3 ========
17:38:13 2021-09-08 15:38:13.989+0000 INFO Initializing system graph model for component 'security-users' with version -1 and status UNINITIALIZED
[...]
17:38:14 2021-09-08 15:38:14.314+0000 INFO Bolt enabled on [0:0:0:0:0:0:0:0%0]:7687.
The log includes an IPv6-formatted address and the default, regex-based, wait strategy will not match.
Activity
bsideup commentedon Sep 8, 2021
ooooh, that's a fun one. We should probably not be that strict about the listener at all, but instead wait for just
Bolt enabled onor some other indicator, thanks for reporting!Would you like to submit a PR with a fix? We will be releasing 1.16.1 relatively soon and will be happy to include it!
fbiville commentedon Sep 8, 2021
Yep, I'll try to send a PR now.
Make Bolt log wait strategy more lenient (#4455)