Hey Yolanda,
Sorry for the delay. Joe pointed me in the right direction. Local state was
left behind in its default path and as such, moving versions caused the
nodeid to be redefined.
I cleared the bogus node IDs, mive state under a consistent path and tested
again and it worked as expected
Andre,
I attempted to reproduce your test case with a three node cluster
(unsecured in my case since I figured it may not be relevant). Upon initial
startup of nodes in 1.1.0, using the linked 1.0.0 conf directory, each node
was disconnected from the cluster until the vote determined which flow
All,
Still testing with the RC but this morning while searching for some API
details I noticed the case of duplicated nodes seems to have been
previously observed:
https://community.hortonworks.com/questions/59685/nifi-cluster-duplicate-nifi-node.html
On Sat, Nov 26, 2016 at 2:44 AM, Joe Witt
Joey - thanks for flagging that. That is indeed a no-go.
I've changed my vote to a -1 (binding) and as the RM I am cancelling this vote.
I'll get another RC together as soon as Joey's PR is available. It
will incorporate that as well as Koji's timeout adjustments for the
site-to-site-client
-1 (non-binding)
While verifying the LICENSE and NOTICE it occurred to me that some test data
that was included (by me sadly) in TestExtractHL7Attributes is MPL (category B)
licensed, which while ok for binary dependencies is not permitted for source
dependencies.
I'll PR and remove these
It's not a blocker for me, as it is clearly a config issue versus an actual
license issue
On Nov 25, 2016 4:18 PM, "Koji Kawamura" wrote:
> Tony, Joe,
>
> Sorry about the nifi-websocket-bundle Rat check issue, I should have
> added apache-rat-plugin exclude configuration
Thanks Koji. It would just end up getting excluded anyway from the
RAT check so that isn't a big deal. The point of RAT is to provide a
tool to help make enforcing licensing and proper notice. Ultimately,
we're responsible to do that regardless of RAT tells us (or doesn't
tell us).
Thanks for
Tony, Joe,
Sorry about the nifi-websocket-bundle Rat check issue, I should have
added apache-rat-plugin exclude configuration in its pom.xml like
other projects such as nifi-toolkit-tls does.
Created a JIRA for that:
https://issues.apache.org/jira/browse/NIFI-3103
I'll send a PR immediately. I
Tony
I don't believe I ram contrib-check on Windows or Linux. I did that on osx.
My win environment is win10 home. Very recent Java 8 amd maven 3.3.9.
Thanks
Joe
On Nov 25, 2016 11:36 AM, "Tony Kurc" wrote:
> Joe Witt,
> I'm not able to build on Windows 10, I'm failing a
No problem. Thanks
On Nov 25, 2016 10:42 AM, "Andre" wrote:
Joe,
The non-binding was more in the sense it is not a show stopper (as I don't
foresee too many people upgrading that way) but I guess I should have made
it more explicit. :-)
Regarding SNAPSHOT, my bad...
Joe,
The non-binding was more in the sense it is not a show stopper (as I don't
foresee too many people upgrading that way) but I guess I should have made
it more explicit. :-)
Regarding SNAPSHOT, my bad... good news is that
nifi.build.revision=1b2b9f1
which happens to be Andy's last commit
Andre
BTW as a member of the PMC your votes are binding.
I am not quite sure the state of your snapshot version relative to the
release version. Definitely worth filing a JIRA and doing further
evaluation.
Thanks
Joe
On Fri, Nov 25, 2016 at 10:13 AM, Andre wrote:
> Joe,
Joe,
-0 (non-binding)
When testing "rolling upgrade" I noticed that as nodes restarted they were
given new Node Ids (I suspect the NodeId is related to the version they
run?). This results on a cluster with 50% of nodes showing up as
disconnected.
Not sure if this is particular to my test
+1 (non-binding)
Full build with contrib-check on OS X and CentOS 7
Classic flows tested on unsecured and secured standalone instance, and on
unsecured cluster with embedded ZK.
Great job team: 265 JIRAs, it's huge!
PS: happy thanksgiving!
2016-11-24 15:27 GMT+01:00 Koji Kawamura
+1 (non-binding)
Ran full clean build w/contrib-check
Built on OSX
Ran a various test on secure/un-secure/cluster/standalone NiFi environments
I really like the updated colorful UI icons!
Thanks!
On Thu, Nov 24, 2016 at 6:04 PM, Andrew Psaltis
wrote:
> +1
+1 (non-binding)
Ran full clean build w/contrib-check
Built on OSX
Ran a variety of use case importing templates, processing data on OSX
Tested using Chrome and Safari.
Lots of great features -- UI seems snappier
Thanks
On Thu, Nov 24, 2016 at 2:04 AM, Joe Witt wrote:
>
To clarify the correct place to find the artifacts is...
The source zip, including signatures, digests, etc. can be found at:
https://dist.apache.org/repos/dist/dev/nifi/1.1.0/
Thanks
Joe
On Thu, Nov 24, 2016 at 2:06 AM, Joe Witt wrote:
> +1 (binding)
>
> Full clean build
+1 (binding)
Full clean build w/contrib-check passes.
Built on OSX, Windows 10, and Linux.
Ran actual usage tests on each osx, linux, and windows.
Tested using Chrome, Firefox, and Edge browsers.
There are a lot of great features, improvements and fixes in this release!
Thanks
Joe
On Thu, Nov
Hello Apache NiFi Community,
I am pleased to be calling this vote for the source release of Apache NiFi,
nifi-1.1.0.
The source zip, including signatures, digests, etc. can be found at:
https://repository.apache.org/content/repositories/orgapachenifi-1092
The Git tag is nifi-1.1.0-RC1
The Git
19 matches
Mail list logo