Use WaitUntilReady retry strategy for test pool

[torcolvin]

Followup to #7642. I don't know why this only hits MasterIntegration and not Sync Gateway Integration Tests but I could reproduce locally.

[Copilot]

Pull Request Overview

This PR updates the cluster agent initialization to use a new CouchbaseClusterWaitUntilReadyOptions struct instead of a simple timeout duration, and adds retry strategy configuration for improved reliability during test execution. This is a follow-up to address integration test issues that were occurring in MasterIntegration tests.

• Introduces CouchbaseClusterWaitUntilReadyOptions struct to encapsulate timeout and retry strategy configuration
• Updates NewClusterAgent function signature to accept the new options struct instead of a simple timeout
• Adds BestEffortRetryStrategy for test pool initialization to improve test reliability

Reviewed Changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 1 comment.

File	Description
base/gocb_utils.go	Defines new options struct and updates NewClusterAgent to use retry strategy configuration
base/main_test_cluster.go	Updates test cluster creation to use new options struct with BestEffortRetryStrategy
rest/x509_test.go	Updates x509 test to use new options struct with retry strategy
rest/server_context.go	Updates server context cluster agent initialization to use new options struct

[Copilot]

The server context initialization doesn't specify a RetryStrategy in CouchbaseClusterWaitUntilReadyOptions, which means it will use the zero value (nil). This is inconsistent with test code that explicitly sets BestEffortRetryStrategy. Consider adding an explicit RetryStrategy for consistency.

Suggested change

	Timeout: 5 * time.Second,
	Timeout: 5 * time.Second,
	RetryStrategy: gocbcore.NewBestEffortRetryStrategy(nil),

[bbrks]

I'm not sure I understand what this change actually does.

Everywhere RetryStrategy is set is nil, either explicitly or implicitly?

[bbrks]

nvm, I see - nil is a parameter...

[torcolvin]

I'm not sure I understand what this change actually does.

Everywhere RetryStrategy is set is nil, either explicitly or implicitly?

The default retry strategy is a fast failure retry strategy, from https://github.com/couchbase/gocbcore/blob/e5f9929c8e4dcd4a4d6dcfda751f0748db71cced/agent.go#L182

It is potentially not correct to have a fast failure retry strategy for the cluster agent. We do not have one for the bucket connections, but it is out of scope to change this for this PR.