Run e2e test in parallel #545
Conversation
chxchx
force-pushed
the
parallelTests
branch
3 times, most recently
from
87b2c52
to
096e692
Compare
|
/assign @sebastienvas |
tests/e2e/util/kubeUtils.go
Outdated
| @@ -106,8 +106,8 @@ func GetIngress(n string) (string, error) { | |||
| func GetIngressPod(n string) (string, error) { | |||
| retry := Retrier{ | |||
| BaseDelay: 5 * time.Second, | |||
| MaxDelay: 30 * time.Minute, | |||
| Retries: 10, | |||
| MaxDelay: 20 * time.Minute, | |||
There was a problem hiding this comment.
Make it 5 minutes. We should not have to wait 20 mn for a test, that must be a mistake.
tests/e2e.sh
Outdated
| @@ -30,21 +30,46 @@ TESTS_TARGETS=($(bazel query 'tests(//tests/e2e/tests/...)')) | |||
| FAILURE_COUNT=0 | |||
| SUMMARY='Tests Summary' | |||
|
|
|||
| cd ${ROOT} | |||
| declare -A pid2testname | |||
| declare -A pid2logfile | |||
|
|
|||
| for T in ${TESTS_TARGETS[@]}; do | |||
There was a problem hiding this comment.
Could you make this configurable ie run the test sequentially by default and run in parallel when a flag is passed.
|
PTAL |
| ;; | ||
| esac | ||
| # Filter -p out as it is not defined in the test framework | ||
| ARGS+=( ${!i} ) |
There was a problem hiding this comment.
Don t you need to have this inside the case ? I am not sure I understand that bash voodoo.
There was a problem hiding this comment.
This is trying to say, append every flag to ARGS except for -p or --parallel. We can no longer use ${@} to have all command line flags to e2e.sh to fall through and apply to the test targets because the new -p flag is meant for the shell script only, and is not even defined in the go programs, so we have to filter it out.
There was a problem hiding this comment.
ok I did not see the continue.
|
@chxchx you might need to rebase with master, also could you please squash your commits into one? |
|
@odeke-em Thanks for the reminder. I would like to keep a complete commit history just so the reviewers do not have to read the entire PR again, but will definitely squash right before merging. |
|
Makes sense, thanks for letting me know @chxchx! |
|
@chxchx you don't need to squash your commits. Github will do this for you. |
|
@sebastienvas Would you like to approve again in Submit-queue way :) |
|
Let's try to get this in first, and I will merge this to my pr and finish that. |
|
/lgtm |
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: chxchx, sebastienvas Assign the PR to them by writing The full list of commands accepted by this bot can be found here.
Needs approval from an approver in each of these OWNERS Files:
You can indicate your approval by writing |
|
/test all [submit-queue is verifying that this PR is safe to merge] |
|
/assign @geeknoid |
|
Automatic merge from submit-queue |
Automatic merge from submit-queue Run e2e test in parallel This experimental PR aims to improve e2e framework to reduce execution time by running tests on mixer and bookinfo concurrently, and yet console logs are collected atomically without interleaving, preserving the abstraction of serial execution. Empirical results have shown about 4-5 min reduction (25% improvement) in time elapsed. **Release note**: ```release-note NONE ``` Former-commit-id: a6aa17e
Automatic merge from submit-queue Run e2e test in parallel This experimental PR aims to improve e2e framework to reduce execution time by running tests on mixer and bookinfo concurrently, and yet console logs are collected atomically without interleaving, preserving the abstraction of serial execution. Empirical results have shown about 4-5 min reduction (25% improvement) in time elapsed. **Release note**: ```release-note NONE ``` Former-commit-id: a6aa17e
Automatic merge from submit-queue Run e2e test in parallel This experimental PR aims to improve e2e framework to reduce execution time by running tests on mixer and bookinfo concurrently, and yet console logs are collected atomically without interleaving, preserving the abstraction of serial execution. Empirical results have shown about 4-5 min reduction (25% improvement) in time elapsed. **Release note**: ```release-note NONE ``` Former-commit-id: a6aa17e
* fix plurality of EnvoyFilter Signed-off-by: Shriram Rajagopalan <shriramr@vmware.com> * more conditions
…ss (istio#255) (istio#545) This is a reimplementation of the following for the Maistra 2.1 / Istio 1.8 rebase: - MAISTRA-1723: Allow disabling NodePort gateway support (istio#180) - Ensure Pilot can run without privileges for reading nodes (istio#125) A new istiod flag is added, `disableNodeAccess`, that will disable any feature relying on access to Node objects. Co-authored-by: Brad Ison <brad.ison@redhat.com>
…ss (istio#255) (istio#545) This is a reimplementation of the following for the Maistra 2.1 / Istio 1.8 rebase: - MAISTRA-1723: Allow disabling NodePort gateway support (istio#180) - Ensure Pilot can run without privileges for reading nodes (istio#125) A new istiod flag is added, `disableNodeAccess`, that will disable any feature relying on access to Node objects. Co-authored-by: Brad Ison <brad.ison@redhat.com>
…ss (istio#255) (istio#545) This is a reimplementation of the following for the Maistra 2.1 / Istio 1.8 rebase: - MAISTRA-1723: Allow disabling NodePort gateway support (istio#180) - Ensure Pilot can run without privileges for reading nodes (istio#125) A new istiod flag is added, `disableNodeAccess`, that will disable any feature relying on access to Node objects. -------- Co-authored-by: Brad Ison <brad.ison@redhat.com> Co-authored-by: Jacek Ewertowski <jewertow@redhat.com> Signed-off-by: Yann Liu <yannliu@redhat.com>
This experimental PR aims to improve e2e framework to reduce execution time by running tests on mixer and bookinfo concurrently, and yet console logs are collected atomically without interleaving, preserving the abstraction of serial execution.
Empirical results have shown about 4-5 min reduction (25% improvement) in time elapsed.
Release note: