When setting up the docker container in this module, there's a an additional script that runs (00-trust-user-ca) that takes some time to execute. This can occasionally take longer than the current 20s timeout, so we are bumping it here
This alias will be used by commands in a test when using the minio client. An alias needs to be defined in order for commands like retrieving an object from the bucket to work successfully.
There have been a few instances where the e2e tests runtime hit this limit. For now, we will update the timeout to make things pass more regularly. There will also be some additional activity to reorganize the e2e tests as runtime has become too long.
* testing/e2e: Use a separate config file for database init
The other config file contains templated parameters, which was causing
errors when loading the config file, even though the values were never
used.
* testing/e2e: add target pagination test
Exercises target pagination both over the CLI and the API. These tests
run in <10s on my machine, though they each involve creating 1000
targets (with current defaults).
---------
Co-authored-by: Michael Li <michael.li@hashicorp.com>
* test(e2e): Add rate limit test (#4135)
(cherry picked from commit 9bf350d622a14ea39feaca29861c738a5deff9bb)
* test(e2e): Modify rate limit test due to timing issues (#4145)
The rate limit test was running into failures due to some hard-coded assumptions (waiting for a command to last longer than 3 seconds to determine if the CLI auto-retried). This change makes it more resilient.
(cherry picked from commit 3535f15922c37fda5457f02cf8ae894203359042)
* test(e2e): Add test for LDAP
This adds the infrastructure and a test for an LDAP workflow
* test(e2e): Refactor test packages
This reduces the number of scenarios needed
* CR
* fixup! test(e2e): Add test for LDAP
* test(e2e): Add test for controller led worker registration
* chore(e2e): Rename log files to ensure uniqueness
This updates the log files to be based on the scenario name instead of just the test package in case some scenarios run the same test package
* chore(e2e): Allow ssh to targets to assist with debugging
* test(e2e): Add test with scp operations
* CR: Use io.CopyN to optimize memory allocation
This mitigates the need to allocate a 256K string
* chore(enos): Move terraform-enos-aws-infra to boundary
The terraform-enos-aws-infra repo was a terraform module to set up the base aws infrastructure. Vault made some local updates that might mitigate flakiness we've been seeing ("can't delete due to dependencies" error). We are replacing our usage of that module Vault's local copy.
* chore(enos): Modify service user to allow DeleteRoute
* fixup! chore(enos): Move terraform-enos-aws-infra to boundary
* chore(enos): Rename variable for clarity
* test(e2e): Add worker health check when setting up infra
* test(e2e): Set limits on postgres container
This was done to potentially reduce some load on runners
* refact(e2e): Add env for workers
* test(e2e): Add test for single-hop worker
* fixup! refact(e2e): Add env for workers
* fixup! test(e2e): Add test for single-hop worker
* fixup! test(e2e): Add test for single-hop worker
* fixup! test(e2e): Add test for single-hop worker
* fixup! test(e2e): Add test for single-hop worker