Thanks for running this all down Joe/Aldrin/Oleg. I think Joe's
suggestion of having a test which checks for this sort of environment
condition and fails with a nice explanation would be really nice.
It looks like Oleg fixed the copyright concerns by removing the lines
which appear to have been u
PROBLEM SOLVED. The problem was a mixture of DNS and routing.
For a couple of reasons, including possibly bad router config, each had a
mixture of a bad nameserver configuration or bad routing to the
nameserver. So, they couldn't resolve their own hostnames to an IP
address. That's not unique t
The OSX system passed the tests switching to a different network. Putting
it back on the problem network and checking KafkaPublisherTest-output.txt I
see repeated occurrences of the following. (fat fingered, ignore any
obvious typos)
17:01:59,225 WARN kafka-producer-network-thread | produce
Well that's great to hear, Joe. Thanks for gathering a bit more info.
We should have tests that are a bit more resilient to varying network
conditions. I know we got nabbed by a few when the project first started
incubation To that end, if you know of anything specific between the two
routers s
Perfect
I was just replying too you again ;). Indeed ‘test’ was not used and I already
commented to that effect in JIRA and removed.
Yes, I was wondering why they fail as this is the first time someone reported
KafkaPublisherTest failures, so kind of curious, so please feed all the info
you can
Oleg,
I'm not familiar enough with Kafka to know what the tests need, but looking
at the test source now I don't think the "String test" on line 70 is used
anywhere, so I think you can just remove it.
The failures where pretty much the same on all 3 systems. But based on
Aldrin's question about
Aldrin,
Wow, great catch on it being a network issue, the OSX system has
successfully finished a "mvn -Pcontrib-check clean install" run!
I haven't had any other problems on this router, but something must be up.
Now the "fun" starts now to figure out what's wrong? ;-)
I'll let you know if it's
Joe
Thanks for pointing that out. I’ll remove the quote from CNN article, just used
it as sample text. Do you see any other issue with sample data?
Also, which systems did you observe failures described below?
Cheers
Oleg
> On Apr 9, 2016, at 2:47 PM, Joe Skora wrote:
>
> -1 (non-binding)
>
Aldrin,
Yes same network, nothing novel that I know of, they all have internet
access and I've disabled firewalls for comparison with no affect. This is
a new router, so I will move one to the old router and see if that makes a
difference.
The validate/valudate mismatch is a fat fingering error
The vote for nifi 0.6.1 is cancelled.
While test issues are one thing copyright issues are another. I looked at
the test Joe highlighted and believe we must remove both the quote and the
apparent news article. Will address and send another rc asap. If someone
could tackle the test failing tests
Joe,
Are these all running on the same network? If so, anything novel or
different about it?
Also, a I assume these are remnants of debugging, but noticed some
differences in terms of the source at the tagged RC and your stacktraces.
TestGetHTTP, for instance, only has 431 lines and the test lis
-1 (non-binding)
Builds and checkouts out ok, but 3 of 4 systems I tried fail in
KafkaPublisherTest unit tests with Java 7 and Java 8. Details and sample
results are included below.
The tests appear to handle ConsumerTimeoutException when it is expected,
but the very first call to ConsumerIterat
+1 (non-binding)
Ran through the helper on OS X and ran a couple templates. Also did a contrib
check build on Windows 8. - - - - - - Joseph
Percivalllinkedin.com/in/Percivalle: joeperciv...@yahoo.com
On Friday, April 8, 2016 5:41 PM, Tony Kurc wrote:
+1 (binding)
On Fri, Apr 8, 2016
+1 (binding)
On Fri, Apr 8, 2016 at 3:02 PM, Matt Burgess wrote:
> +1 (non-binding)
>
> Ran through the release verification process (checksums, hashes, etc.), ran
> tests and some example flows, everything looks good.
>
> Regards,
> Matt
>
> On Wed, Apr 6, 2016 at 5:33 PM, Joe Witt wrote:
>
>
+1 (non-binding)
Ran through the release verification process (checksums, hashes, etc.), ran
tests and some example flows, everything looks good.
Regards,
Matt
On Wed, Apr 6, 2016 at 5:33 PM, Joe Witt wrote:
> Hello Apache NiFi Community,
>
> I am pleased to be calling this vote for the source
+1 (binding)
Verified signature, hashes, etc. Set up the application with various
configurations (standalone/clustered) (secured/unsecured) and all ran as
expected.
Matt
On Wed, Apr 6, 2016 at 5:33 PM, Joe Witt wrote:
> Hello Apache NiFi Community,
>
> I am pleased to be calling this vote for
+1 (binding)
Ran through release helper and everything checked out.
Ran example flows and functioned as expected.
On Wed, Apr 6, 2016 at 11:16 PM, Aldrin Piri wrote:
> +1, binding
>
> Build and tests worked on OS X and Windows 7. Branch looked good on our
> Travis build and ran some flow templ
+1, binding
Build and tests worked on OS X and Windows 7. Branch looked good on our
Travis build and ran some flow templates from reviews from some of the
associated issues with anticipated results.
Verified hashes and signatures.
On Wed, Apr 6, 2016 at 5:33 PM, Joe Witt wrote:
> Hello Apache
Hello Apache NiFi Community,
I am pleased to be calling this vote for the source release of Apache
NiFi 0.6.1.
The source zip, including signatures, digests, etc. can be found at:
https://dist.apache.org/repos/dist/dev/nifi/nifi-0.6.1/
The Git tag is nifi-0.6.1-RC1
The Git commit hash is d51b24
19 matches
Mail list logo