Tony,
I tend to agree with your view but I was under the impression the proposed
change does not make changes to NiFi itself?
Correct me if I am wrong - maven still feels quite cryptic to me - but my
understanding was that the PR only configures tests to be ran with a
particular variable java
Andre,
I've been reluctant to do a software change for something that is,
debatably an environment issue. If DISPLAY is set, I would think one would
be reasonably assured one had an x server.
On Mon, Aug 29, 2016 at 8:02 PM, Andre wrote:
> Tony,
>
> Yes. That happens when
Tony,
Yes. That happens when the JDK is not headless. I submitted a PR that
forces maven to run in headless mode which also solves the issue.
Have you had a chance to check https://github.com/apache/nifi/pull/961 ?
Cheers
On Tue, Aug 30, 2016 at 9:57 AM, Tony Kurc wrote:
>
Andre, on the X11 issue, it can occur if you have the DISPLAY environment
variable set, but no X server.
I reported it a while back with what got me past it:
https://issues.apache.org/jira/browse/NIFI-1154
On Aug 29, 2016 11:00 AM, "Yolanda Davis" wrote:
> Resending
Builds with JDK 1.8.0_60 so I am a +0 in total. It's been a long time
(maybe 0.5?) since I attempted a full nifi build rather than just building
the processors I am working with. It seems like some time in that window a
JDK incompatibility was introduced.
On Mon, Aug 29, 2016 at 2:58 PM, Alan
-1 (non-binding) from me, but likely an issue on my machine.
I get a compile error in nifi-framework-cluster on this release and master
using an empty maven repository.
Mac OS X 10.10.5. Maven 3.2.5. Oracle Java 1.8.0_31
The end of maven debug output (which doesn't even say which file is
Still voting a +1 for the release, but I recently created two JIRAs for the
issues I encountered while testing nifi-1.0.0-RC1:
https://issues.apache.org/jira/browse/NIFI-2695 - Access Denied messages
should include more information
https://issues.apache.org/jira/browse/NIFI-2699 - Improve
+1 (binding)
Verified keys, hashes, build, etc. Ran standalone and cluster secured
instances.
Evaluated the issues Koji reported and created two minor JIRAs [1] [2] to
perform some additional clean up in the next release.
Thanks, Joe P!
Matt
[1] https://issues.apache.org/jira/browse/NIFI-2694
+1 (binding)
Verified the keys & md5 hash. Was able to startup the node and verify things
are behaving as expected.
Verified that the README, NOTICE, and LICENSE files exist and appear to be
correct.
Thanks for pulling this together, Joe P!
> On Aug 26, 2016, at 12:25 PM, Joe Percivall
+1 (non-binding) Release this package as nifi-1.0.0
- Verified checksums, signature, LICENSE/NOTICE/README
- Built source with contrib-check
- Tested standalone and clustered instances
There were two issues that I ran into that are minor and are
user-experience-based. The first issue
+1 (non-binding)
- Verified checksums, signature, and commit hash
- Ran build on OS X with all tests and contrib-check
- Tested recent ExtractHL7Attributes (NIFI-2564) and TransformXml (NIFI-2142)
improvements, and template import/export on an unsecured cluster with external
ZK
On Aug 26,
+1, binding
Hashes check out
Signature looks good
Build looked good and ran on OS X, CentOS, Debian
Sample flows from templates check out and look good.
Only small hangup was none of the TestListenSMTP tests passing in Windows
10 as created but worked with relaxed timeouts. Created a JIRA issue
+1 (binding) Release this package as nifi-1.0.0
- Verified checksums, signature, commit hash, and LICENSE/NOTICE/README
- Performed full build with contrib-check
- Tested a standalone unsecure instance and then used the toolkit to create
a secure cluster
On Mon, Aug 29, 2016 at 11:00 AM, Yolanda
Resending on appropriate thread:
+1 (non-binding)
Ran full test suite, with rat and contrib-check. Verified, readme, notice,
license, commit ID, hashes and checksums. Tested flows and templates in
standalone cluster (on Mac OS).
On Fri, Aug 26, 2016 at 12:25 PM, Joe Percivall <
+1 (non-binding)
Verified checksums, GPG signature, commit hash, LICENSE/README/NOTICE,
etc. Tested various flows, templates, controller services on unsecure
standalone and secure 3-node cluster, looking good!
On Fri, Aug 26, 2016 at 12:25 PM, Joe Percivall
@Pierre, if you don't want it I'd turn off updates, I did NOT get any
special notice until after it had installed.
On Mon, Aug 29, 2016 at 8:09 AM, Pierre Villard wrote:
> @Joe I don't have the anniversary update yet.
>
> 2016-08-29 13:56 GMT+02:00 Joe Skora
@Joe I don't have the anniversary update yet.
2016-08-29 13:56 GMT+02:00 Joe Skora :
> To followup, the TestListenSMTP timeout problems got much better when I
> changed a questionable DNS setting on the router. They still won't pass
> within the 5 seconds currently configured
Pierre and Andre,
Thanks for reporting the issue where I forgot that some European and Australian
time zones are represented as 4 characters. Obviously the test is overly
specific with the regex, as the purpose is to ensure that the first serialized
line is the date and time, and that is still
+1 (non-binding). I ran through the release helper, including hashes,
build, tests. No showstopping issues for me.
James
On Fri, Aug 26, 2016 at 9:25 AM, Joe Percivall <
joeperciv...@yahoo.com.invalid> wrote:
> Hello Apache NiFi Community,
>
> I am pleased to be calling this vote for the
Full clean build w/contrib check good.
Ran standalone NiFi and secured standalone.
-0.9 (non-binding)
Because I found two issues during the test.
1. Confusing log message with tls-toolkit.sh "No hostnames specified"
This is a trivial one, and this is NOT the reason why I did negative
fraction
+1 (non-binding)
Full build with contrib-check on Mac OSX, Windows 10 and CentOS 7.
I had the same issue as reported by Joe on SMTP tests with Windows 10
build. Workaround with timeouts works.
I also reported two minor issues related to timezones (NIFI-2682) / locale
(NIFI-2683) aspects.
created NIFI-2688 and PR963(master) and PR964 (RC1) to address the regex
issue.
On Mon, Aug 29, 2016 at 3:07 PM, Andre wrote:
> All,
>
> Had a look in the groovy code and reason is quite obvious:
>
> System locale does not match what the code expects via regex.
>
>
>
All,
Had a look in the groovy code and reason is quite obvious:
System locale does not match what the code expects via regex.
So, for the record, here are the test times.
TestListenSMTP.validateSuccessfulInteraction - 59s 74ms
TestListenSMTP.validateSuccessfulInteractionWithTls - 56s 111ms
TestListenSMTP.validateTooLargeMessage - 19s 123ms
This system runs Windows 10 Pro, Version 1607, OS Build 14393.82 on an
Change that to +1 (non-binding)
With longer timeouts the TestListenSMTP tests are now passing.
On Sun, Aug 28, 2016 at 10:04 PM, Joe Skora wrote:
> -1 (non-binding)
>
> It looks good in general, but Windows 10 unit tests consistently fail in
> TestListenSMTP.
>
> It builds
Ok, I'm changing my vote to +1!
I don't know the root cause is, but I changed the "latch.await" timeouts
from 5 to 60 seconds and the test pass. :-D
This was the only system that built the 1.0.0-beta on the first try, the
only change since then is the Windows 10 Anniversary update last week.
Ah sorry, misread your message. I see you ruled out an EOL issue.
Any idea what would change in regards to the SMTP tests in Windows 8 vs 10 (or
your system vs mine)?
Joe
- - - - - -
Joseph Percivall
linkedin.com/in/Percivall
e: joeperciv...@yahoo.com
On Sunday, August 28, 2016 10:10 PM,
JoeS,
I tested the build on Windows 8 and everything worked fine for me. Wouldn't an
EOL issue show up in all Windows builds?
Joe
- - - - - -
Joseph Percivall
linkedin.com/in/Percivall
e: joeperciv...@yahoo.com
On Sunday, August 28, 2016 10:05 PM, Joe Skora wrote:
-1
-1 (non-binding)
It looks good in general, but Windows 10 unit tests consistently fail in
TestListenSMTP.
It builds fine on OS X and Ubuntu has one error described below.
It runs fine on OS X and Ubuntu including AWS S3 and HandleHTTP processors.
1. signature verifies
2. hashes verify
3. build
Verified all checksums and GPG signature; verified contrib check and RAT
license check; verified git commit ID, checked the LICENSE, README, and NOTICE
files, and ran the binary and performed the usual actions within the UI.
I then used the encrypt-config toolkit to encrypt the banner UI text
Downloaded source
Verified checksums
Ran build with contrib-check and empty local repo
Ran build with –T2 concurrency and empty local repo
Used tls-toolkit to generate 4 localhost keystore, truststore,
nifi.properties and a single client certificate
Started all 4 instances verified no port
+1 (binding)
Full build starting with empty maven repository on Ubuntu 14.04 (x86_64)
successfully.
Verified signatures and hashes. Reviewed LICENSE/NOTICE.
Ran a simple flow with no issues. Ran a few nifi-toolkit commands.
On Sun, Aug 28, 2016 at 5:29 PM, Joe Witt
+1 (binding)
Full clean build w/contrib check good.
Verified signatures (and strength), hashes, and LICENSE/NOTICE within
sources and convenience binaries.
Verified the commit the release was based on.
Ran single and multi-node clusters with favorable results.
For future release please do
33 matches
Mail list logo