Re: [ANNOUNCEMENT] Nexmark included to the CI

Hi Etienne, It's awesome for working on these useful dashboards. I am getting TPC-H benchmark running on Flink and Dataflow Runner. I could work on similar dashboards for TPC benchmark after code merged. Also, it's great to have a dashboards for Dataflow. Best, Kai ᐧ On Wed, Jul 11, 2018 at

Re: Automatically create JIRA tickets for failing post-commit tests

I think this looks good, we should enable the plugin and try it out. Concrete details of the follow-up tasks (auto-assignment, triage, and dashboarding) will probably depend on how functional the plugin is and what the test failures data looks like. Regards, Anton On Wed, Jul 11, 2018 at 5:00 PM

Re: Vendoring / Shading Protobuf and gRPC

Got a fix[1] for Andrews issue which turned out to be a release blocker since it broke performing the release. Also fixed several minor things like javadoc that were wrong with the release. Solving it allowed me to do the publishing in parallel and cut the release time from 20+ mins to 8 mins on

Re: Running post-commit tests on every PR

Thanks Mikhail. I'd like to close this discussion now. The idea of running post commits frequently is to avoid post commits tests failing ASAP. However, bugs are supposed to be caught in precommit tests. With the advantage of fast running time of precommit tests, improving precommit tests is a

Re: Automatically create JIRA tickets for failing post-commit tests

@Yifan Zou I believe that we should test-drive the system with tickets + PR first and decide on email notification later. We already have tests failure emails sent to commits@, I believe most people filter out or not signed up for that list though. It creates only one ticket, and keeps it for

Re: Automatically create JIRA tickets for failing post-commit tests

+1 to Andrew's concerns. Leaving the tickets unassigned will cause the ticket being ignored and no actions being taken. I can see the challenges on ticket assignment. Like Mikhail mentioned, the plugin does not support dynamic assignments. We have to implement custom script to determine the

Re: Automatically create JIRA tickets for failing post-commit tests

That sounds great. You should add this detail to the doc. On Wed, Jul 11, 2018 at 4:29 PM Mikhail Gryzykhin wrote: > We already have component for this purpose: "test-failures". All tickets > created will go to that component. As an option, we can add link to view > list of open JIRA tickets to

Running post-commit tests on every PR

Hello everyone, Rui suggested to run post-commit tests on every pull request. Lets utilize this forked thread to discuss this suggestion. Best regards, --Mikhail On Wed, Jul 11, 2018 at 4:25 PM Mikhail Gryzykhin wrote: > Hi Rui, > > I would suggest to start another thread for that

Re: Automatically create JIRA tickets for failing post-commit tests

We already have component for this purpose: "test-failures". All tickets created will go to that component. As an option, we can add link to view list of open JIRA tickets to PR template. We also would want to create graph on dashboard with amount of unassigned and assigned bugs. I believe that

Re: Automatically create JIRA tickets for failing post-commit tests

Hi Rui, I would suggest to start another thread for that discussion. Lets keep this discussion focused on automated JIRA tickets creation. --Mikhail On Wed, Jul 11, 2018 at 4:12 PM Rui Wang wrote: > Maybe this is also a good thread to start the discussion that if we want > to enforce

Re: Automatically create JIRA tickets for failing post-commit tests

So it sounds like you will want to create a component for untriaged issues so they are easy to find. I like the idea of distributing the work of triaging post commit failures to new PR authors as a condition of merging. I feel like we will just be filling JIRA with spam if the issues are

Re: Automatically create JIRA tickets for failing post-commit tests

Maybe this is also a good thread to start the discussion that if we want to enforce postcommit test for every PR. Can we afford the cost of longer waiting time to catch potential bugs? -Rui On Wed, Jul 11, 2018 at 4:04 PM Mikhail Gryzykhin wrote: > That's a valid point. > > Unfortunately, the

Re: Automatically create JIRA tickets for failing post-commit tests

That's a valid point. Unfortunately, the JiraTestResultReporter plugin does not have features to dynamically assign owners. Additionally, I don't think it is always easy to find proper owner for post-commit tests at first glance, since they usually cover broad specter of issues. My assumption is

Re: Vendoring / Shading Protobuf and gRPC

We discussed this in person, sounds like my issue is known and will be fixed shortly. I'm running builds with '-Ppublishing' because I need to generate release artifacts for bundling the Beam SQL shell with the Google Cloud SDK. Hope to eventually just use the Beam release, but we are currently

Re: Automatically create JIRA tickets for failing post-commit tests

Hi Mikhail, I like the proposal! Hopefully this can replace the constant stream of build failure emails. I noticed one detail seems to be missing: How will new issues be assigned to the proper owner? Will the tool do this automatically or will we need someone to triage new issues? Andrew On

Automatically create JIRA tickets for failing post-commit tests

Hi everyone, I want to add an automatic JIRA tickets creation for failing post-commit tests. I wrote up design proposal doc with more details on this: https://docs.google.com/document/d/1kpsTy0sYJkLqlZvkPalDkqzBbpu-Wug0z-oWIVPo6UI Quick summary: I suggest to utilize JiraTestResultReporter

Re: Vendoring / Shading Protobuf and gRPC

Andrew, to my knowledge it seems as though your running into BEAM-4744, is there a reason you need to specify -Ppublishing? No particular reason to using ByteString within ByteKey and TextSource. Note that we currently do shade away protobuf in sdks/java/core so we could either migrate to using a

Re: [PROPOSAL] Prepare Beam 2.6.0 release

+1 Thanks for volunteering, Pablo On Wed, Jul 11, 2018 at 11:49 AM Jason Kuster wrote: > +1 sounds great > > On Wed, Jul 11, 2018 at 11:06 AM Thomas Weise wrote: > >> +1 >> >> Thanks for volunteering, Pablo! >> >> On Mon, Jul 9, 2018 at 9:56 PM Jean-Baptiste Onofré >> wrote: >> >>> +1 >>> >>>

Re: Broken seed job

The outage likely fixed it. I would have expected running the standalone job with the trigger phrase should have fixed it. Did it not? On Wed, Jul 11, 2018 at 8:45 AM Lukasz Cwik wrote: > I believe there was an outage a few hours ago. > > On Wed, Jul 11, 2018 at 8:36 AM Łukasz Gajowy > wrote:

Re: [PROPOSAL] Prepare Beam 2.6.0 release

+1 Thanks for volunteering, Pablo! On Mon, Jul 9, 2018 at 9:56 PM Jean-Baptiste Onofré wrote: > +1 > > I planned to send the proposal as well ;) > > Regards > JB > > On 09/07/2018 23:16, Pablo Estrada wrote: > > Hello everyone! > > > > As per the previously agreed-upon schedule for Beam

Proposal: Apache Beam Contributor Metrics: Collection, Display, Actions

I am proposing an html dashboard for some key contributor metrics, initially the weekly passing rate of post-commit tests and the speed of pre-commit tests, then expanding to code review and release metrics. https://s.apache.org/beam-contributor-metrics Comments welcome. Alan

Re: Broken seed job

It's totally fine. The problem is now gone even though I have taken no action to fix it. I suppose Jenkins was restarted? BTW: Is it restarted only on demand by those who have access, or it's done once in a while (periodically)? śr., 11 lip 2018 o 17:07 Lukasz Cwik napisał(a): > Ah, sorry for

Re: Broken seed job

Ah, sorry for my confusion. On Tue, Jul 10, 2018 at 4:59 PM Łukasz Gajowy wrote: > I didn't edit the "Standalone Seed job", only the "SeedJob". Now every > time someone tries to run the seed job ("Run seed job") it results in an > error even despite prior running the standalone job from master

Re: The full list of proposals / prototype documents

Thank you for this link, Etienne. I agree that it doesn’t fit well for design documents page. So, I think it makes sense to add either on wiki or as a part of Nexmark documentation on web site: https://beam.apache.org/documentation/sdks/java/nexmark/

Re: [ANNOUNCEMENT] Nexmark included to the CI

First catch of the nexmark-CI:It seems that there was a change in the direct runner. Query3 (exercise state and timers) - output size should be constant but has increased today => Was there a change in state and timer related code?- the output size of this query is different between batch and

Re: [ANNOUNCEMENT] Nexmark included to the CI

Is someone interested in creating the scripts and dashboards for the other runners? They can be created by copying the existing scripts and dashboards and changing one gradle parameter in the scripts and the table name in the dashboards. I have created the

[ANNOUNCEMENT] Nexmark included to the CI

Hi guys, I'm glad to announce that the CI of Beam has much improved ! Indeed Nexmark is now included in the perfkit dashboards. At each commit on master, nexmark suites are run and plots are created on the graphs. I've created 2 kind of dashboards: - one for performances (run times of the

Build failed in Jenkins: beam_SeedJob #2197

See -- GitHub pull request #5927 of commit cd1326b72617ee1ed40e11ce1ef85b10a61a9123, no merge conflicts. Setting status of cd1326b72617ee1ed40e11ce1ef85b10a61a9123 to PENDING with url

Re: The full list of proposals / prototype documents

Alexey, One doc that can be interesting that I forgot to point out is https://docs.google.com/document/d/1VgnGiVu8vSfm7Et-xAtQYv 0PlEpqeyfmhpQUNPmWRJs/edit?usp=sharingIt is the doc I wrote when I submitted Nexmark PR to ease the reading of the code. It is not a design doc, I don't know if it

Build failed in Jenkins: beam_Release_Gradle_NightlySnapshot #97

See Changes: [robertwb] object() takes no parameters [Pablo] Improving signing of all published artifacts and adding test publication [Pablo] Removing extra publication [cclauss] [BEAM-3761]

Jenkins build is back to normal : beam_SeedJob #2194

See