+1
On Sat, Sep 8, 2018 at 2:21 AM, Matteo Merli wrote:
> Please vote on the proposal for Pulsar graduation to TLP to submit to
> the Incubator PMC.
>
> This vote will stay open for at least 72 hours.
>
>
>
>
> Establish the Apache Pulsar Project
>
> WHEREA
That said, it is stored in Json so it wouldn't be much work to sort
one of the values as a list of strings.
-Ivan
On Thu, Sep 6, 2018 at 10:55 AM, Ivan Kelly wrote:
> PIP updated. The one thing that isn't 100% clean with the dynamic
> configuration keys is that they only support
PIP updated. The one thing that isn't 100% clean with the dynamic
configuration keys is that they only support string as value, so the
list of revoked keys needs some sort of serialization.
-Ivan
On Thu, Sep 6, 2018 at 10:41 AM, Ivan Kelly wrote:
> On Tue, Aug 21, 2018 at 3:57 A
On Tue, Aug 21, 2018 at 3:57 AM, Rajan Dhabalia wrote:
>> I will consider both this and system topics and update the PIP.
>
> I think we should avoid more complexity by introducing system topic to
> store "subject key" and then process it. Instead we can store it to
> global/config zk as it requir
ubmitted to the wiki and
>> discussed in the mailing list.
>>
>> PIP 23: Message Tracing By Interceptors
>> PIP 22: Pulsar Dead Letter Topic
>> PIP 21: Pulsar Edge Component
>> PIP 20: Mechanism to revoke TLS authentication
>> PIP 19: Pulsar SQL
>>
>>
>> How would you assess the podling's maturity?
>> Please feel free to add your own commentary.
>>
>> [ ] Initial setup
>> [ ] Working towards first release
>> [ ] Community building
>> [X] Nearing graduation
>> [ ] Other:
>>
>> Date of last release:
>> 2018-08-02, 2.1.0-incubating
>>
>> When were the last committers or PPMC members elected?
>>
>> 2018-06-11 - Ivan Kelly
>> 2018-06-11 - Jia Zhai
>>
>>
>> --
>> Matteo Merli
>>
>>
that should be +1
On Thu, Sep 6, 2018 at 10:38 AM, Ivan Kelly wrote:
> lgtm +!
>
> On Thu, Sep 6, 2018 at 9:52 AM, Sijie Guo wrote:
>> +1 looks good
>>
>> - Sijie
>>
>> On Wed, Sep 5, 2018 at 5:36 PM Matteo Merli wrote:
>>
>>> Sorry
> The pain comes from issues marked for a bug release milestone but it is
> merged directly to master and never cherry-picked to branch for bugfix
> releases.
> So the release manager has to compare branches and master and also have to
> compare labels, milestones and even look into the details of
+1
I wish to remain part of PMC.
Regarding BookKeeper bylaws, BookKeeper has them because ZooKeeper had
them when we branched off, and my understanding is that zookeeper had
them, because hadoop had them when they branched off. As far as I'm
concerned, they're unnecessary given the default asf by
> All the bug fixes are deferred to release manager cherry-picking to bug fix
> branches. Since master and branches are quickly diverged with changes,
> it is very painful for cherry-picks later and there is no clear process for
> cherry-pick and labelling.
What is painful here? If it's the confli
include it.
>
>
>
>
> On Tue, Aug 28, 2018 at 5:28 AM Ivan Kelly wrote:
>
>> Hi folks,
>>
>> I accidently pushed a change to branch-1.22 to fix #2401.
>>
>>
>> https://github.com/apache/incubator-pulsar/commit/92b809784f631bc6b72cc1116761e8fac02ef41c
Hi folks,
I accidently pushed a change to branch-1.22 to fix #2401.
https://github.com/apache/incubator-pulsar/commit/92b809784f631bc6b72cc1116761e8fac02ef41c
Should I revert immediately? It's a pretty straightforward change
given what the bug was.
-Ivan
> Here, we would like to configure "subject-key-identifier" at every broker
in the cluster dynamically. We also want to perform certain actions once
this configuration-value has been changed.
Ah, i didn't know this existed. Very useful. I will consider both this
and system topics and update the PI
> Instead of using zookeeper, can we consider using a managed ledger or a
system topic for keeping all these revoked keys?
Do we already have an example of a system topic in use? Is there a
custom namespace that's restricted to the admin role?
-Ivan
Hi folks,
This is a PIP to add a mechanism to block TLS client certs from
accessing Pulsar if they have been compromised.
This is a relatively small change, but I thought it best to put it to
the community before moving ahead with it, as people may have opinions
on the approach.
The PIP is here:
The git error can be fixed by adding
false
It already has failOnUnableToExtractRepoInfo as false, which I assumed
would cover this case.
-Ivan
On Mon, Jul 23, 2018 at 6:05 AM, Sijie Guo wrote:
> any more review comments?
>
> - Sijie
>
> On Fri, Jul 20, 2018 at 1:13 PM Sijie Guo wrote:
>
>> Oka
The second change is https://github.com/apache/incubator-pulsar/pull/2177
-Ivan
On Tue, Jul 17, 2018 at 12:45 AM, Sijie Guo wrote:
> Cool. I will wait for these two issues to be merged.
>
> - Sijie
>
> On Mon, Jul 16, 2018 at 3:25 PM Ivan Kelly wrote:
>
>> I
se candidate number that I
> stopped last time.
>
> If you have any questions regarding 2.1 release, feel free to ping me here.
>
> - Sijie
>
>
>
> On Mon, Jul 9, 2018 at 2:44 AM Ivan Kelly wrote:
>
>> It would be good to get tls proxy auth fixes into 2.1 s
It would be good to get tls proxy auth fixes into 2.1 since there's
time to do so now. Without these changes auth + tls do not work if
using a proxy.
-Ivan
On Thu, Jul 5, 2018 at 10:13 PM, Sijie Guo wrote:
> Hi all,
>
> Due the connector packaging issue, we delayed 2.1 release on fixing that.
>
+1
On Thu, Jun 21, 2018 at 2:06 AM, Matteo Merli wrote:
> Following the previous discussion at
> https://lists.apache.org/thread.html/fe60c21b7cdca00918fe25e7ddf9772677342063b9da24c9b84b3329@%3Cdev.pulsar.apache.org%3E
> I am calling a formal vote for keep using "Apache Pulsar" name for the
> pro
> We do have access to the docker client via test containers.
>
> There is no binding between test containers and any test framework all the
> underlying methods and classes and fully exposed so they can be mapped to
> any test harness.
In that case, the first thing to do is try to migrate the smo
Thanks for starting this conversation Ali. The current tests use
arquillian, but there's nothing in the tests that's intimately tied to
it, so if we were to remove it there wouldn't be that much that needs
to change.
On Thu, Jun 14, 2018 at 8:21 PM, Ali Ahmed wrote:
> I am proposing we move our c
>> Two concurrent builds on the same machine will have different
>> workspaces (@2 or @3 will be appended to $WORKSPACE).
>
> I am not sure about that. If taking a look at the jenkins job output, the
> workspace is "
> /home/jenkins/jenkins-slave/workspace/pulsar_precommit_integrationtests"
It onl
> 1) the job is using `$WORKSPACE/.repository` as the maven local repo.
> 2) concurrent build is enabled
> 3) it is a freestyle build, including two maven runs.
Two concurrent builds on the same machine will have different
workspaces (@2 or @3 will be appended to $WORKSPACE). We should
disable con
Hi folks,
I've just come across something very strange in a jenkins build that
others should be aware of in case they hit across the same thing.
https://builds.apache.org/job/pulsar_precommit_integrationtests/1782/
This build is a bulld of the branch
69b5976d8b40789f7f139185ffc1e5cc0bbe4f4b
http
+1 (non-binding)
- Signatures and hashes are good
- LICENSEs and NOTICEs are good
- Rat and tests are clean (some env issues locally, but fine on a clean VM)
- nit: the source distribution is missing the dockerfile and scripts
to build the testing docker image, so I couldn't run integration tests
> Created PR at https://github.com/apache/incubator-pulsar/pull/1748
I'll take a look, but it may be tomorrow before I get the time.
> In some cases, the feedback for the NOTICE file has been to avoid inclusion
> on what was not strictly required, in particular, regarding non ASL
> licensed proje
-1
There's quite a few issues with LICENSE and NOTICE
https://gist.githubusercontent.com/ivankelly/a93f06d2a3075575c43a0dd44165bac1/raw/83551daf958432f532ee53cccaa6397f0c1af7bf/pulsar-libs.org
Some stuff is missing completely. Others are incomplete.
Note that this only applies to the binary dis
On Thu, Apr 19, 2018 at 10:07 AM, Dave Fisher wrote:
> Is it possible to have 2.0.0alpha on a Maven Central release?>
Zookeeper do it all the time.
http://search.maven.org/#artifactdetails%7Corg.apache.zookeeper%7Czookeeper%7C3.5.3-beta%7Cpom
-Ivan
Why not call it 2.0.0Alpha1, or 2.0.0Beta1? Calling it RC will be
confusing for people who haven't read this conversation. The only
problem with alpha or beta is that you won't get a 2.0.0 release in
the end, but a 2.0.1. Many people don't trust '.0' releases anyhow
though.
-Ivan
On Thu, Apr 19,
> It will keep the discussion public, convenient, archived, maintain history,
> and will be mailed out (with github notifications).
>
> What do you all say?
+1 for this. This is how BK does it, and I've found it much easier to use.
-Ivan
Hi folks,
This proposal is to add tiered storage to pulsar.
Storing backlogs on bookies for a long time can get expensive. If
there are other cheaper forms of storage available (S3/HDFS), capex
can be reduced by moving older data to this storage.
The proposed design proposes an interface to do t
> Keep in mind the next Pulsar release needs to be based on a particular
> BookKeeper release. An Apache project cannot release against another
> project's SNAPSHOT since that has not been validated by their release process.
The bookkeeper 4.7.0 release process should be starting soon, so it
sho
t through. There were 246
>> > commits to merge into a codebase that slightly changed in a 4 years
>> > timespan. For the curious, this is the spreadsheet we used to track the
>> > merging.
>> > https://docs.google.com/spreadsheets/d/1jAy3EfjViqNEKpCKpWiRv-
>>
Mon, Feb 26, 2018 at 9:53 PM, Ivan Kelly wrote:
> Hi folks,
>
> I'd like to do some tests with batching and compaction. So far the
> only way I've found to control batching is by setting
> BatchingMaxMessages and MaxPublishingDelay, but this doesn't seem 100%
> det
Hi folks,
I'd like to do some tests with batching and compaction. So far the
only way I've found to control batching is by setting
BatchingMaxMessages and MaxPublishingDelay, but this doesn't seem 100%
deterministic to me.
Does anyone have a better way to control this, or would it make more
sense
Hi folks,
I've just pushed the first[1] in a series of pull requests to add
integration testing to pulsar.
The tests will run with a framework called Arquillian Cube[2], which
can be thought of as a way to run docker-compose within testng.
A lot of the setup and code I'll be pushing comes from B
Couldn't the schema for a topic be stored in another topic with
infinite retention?
So if we have a topic /my-prop/my-namespace/topic-foo, we could have a
topic /my-prop/my-namespace/topic-foo/schema which contains the
different versions of the schema for the original? To get the schema,
you just
On Wed, Jan 10, 2018 at 7:31 PM, Joe F wrote:
> While we are at it, can we also think of "persistent://"? PIP-11 covers
> this only for limited use cases.
>
> Maybe shorten it to p://? That will be 9MB for a million topics, in ZK
> strings (more with utf8).
You could even get rid of the //. http
_
>> > 差出人: Dave Fisher
>> > 送信日時: 2017年12月12日 8:49:13
>> > 宛先: dev@pulsar.incubator.apache.org
>> > 件名: Re: [DISCUSS] Topic Compaction for Pulsar
>> >
>> > Any thoughts Pulsar devs?
>> >
>> > Sent f
Any comments/feedback on this?
-Ivan
On Mon, Dec 4, 2017 at 11:54 PM, Ivan Kelly wrote:
> Hi all,
>
> I've created PIP to add topic compaction to Pulsar. The PIP only
> discusses the compaction mechanism and serving from a compacted topic.
> I've left compaction sche
e Fisher wrote:
>
>> Hi -
>>
>> If you build from source do you have this same issue?
>>
>> Where was your discussion with Matteo? If not on the mailing list then it
>> ought to be for full exposure and informed decision making.
>>
>> Regards,
>> D
-1 on this release.
There's a bunch of netty jars being shipped with the binary
distribution which makes publication fail with an error such as:
2017-12-05 15:31:07,922 - WARN - [pulsar-client-io-2-1:ClientCnx@156]
- [id: 0xd1059600, L:/10.142.0.5:34890 !
R:pulsar3-us-east.c.ivan-184610.internal/
Hi all,
I've created PIP to add topic compaction to Pulsar. The PIP only
discusses the compaction mechanism and serving from a compacted topic.
I've left compaction scheduling to be solved later. The proposed
solution should be kicked off manually.
The master issue is: https://github.com/apache/i
+1 lgtm
Tests done (debian 9):
- sha1, md5, gpg
- apache rat
- unit tests
- NOTICE & LICENSE
- setup cross colo install, tested a few global topics, tested some
local topics, all work well.
Good work!
-Ivan
On Sun, Oct 1, 2017 at 10:55 PM, Joe F wrote:
> This is the first release candidate for
44 matches
Mail list logo