Hi Jerry,
How about checking your release script for uima-as into svn as part of the
uima-as project?
Benefits- it won't get lost, and it's available to others who might someday need
to release uima-as.
Cheers. -Marshall
On 1/7/2018 6:27 PM, Jaroslaw Cwiklik wrote:
> Damn, I now remember
Cancelling due to the problem reported by Eddie Epstein.
Will fix examples and redo the RC
Thanks to all who spent time testing!
Jerry
Damn, I now remember this exact issue biting me before. I'd modified my old
release script but
unfortunately lost it when my disk drive died. Will update the new one to
make sure this is always
tested with every release.
Jerry
On Sat, Jan 6, 2018 at 6:55 PM, Eddie Epstein
- downloaded binary tarball and checked md5sum
- unpacked, set UIMA_HOME, updated PATH and ran adjustExamplePaths.sh
- startBroker.sh worked normally
- ran basic test (MeetingDetectorTaeQueue) from README
- confirmed -log and -perf options on the basic test are working
(although error message
Just to make this more clear, Eddie and I together determined that the AMQ
connection losses where due to
long GCs in the service. This conclusion was based on evidence found in the
service log and correlation with
jConsole stats.
Jerry
On Fri, Jan 5, 2018 at 10:56 AM, Jaroslaw Cwiklik
After reviewing the service log and looking at jConsole there are long GCs
which cause connection loss and recovery.
Looks like AE issue. Strange that these GCs happen if the service is idle.
Someone should profile this AE to confirm
where this and possibly fix the source of excessive garbage.
- Checked signatures - OK
- Tested binary with a dozen CASes - OK
- Built from source & checked signatures - OK
- Tested source build with a dozen CASes - OK
- Spot checked README, RELEASE_NOTES, LICENSE, NOTICE, JIRAs, docs - OK
- Tested TargetServiceId - OK
Minor complaint - The (undocumented)
"Many other" = "all services " were kaput? If so, it could be the broker
failure (what version?) or something else. Sorry cant pinpoint this stuff
with accuracy since the broken connection can be symptom of many problems.
Is this system in the kaput state still?
Jerry
On Tue, Jan 2, 2018 at
There were no long GCs for this service. Note (from above) that many other
services running older versions of uima-as were also showing the same
behavior.
Eddie
On Tue, Jan 2, 2018 at 9:21 AM, Jaroslaw Cwiklik wrote:
> Looks like the client lost connection to the broker for
Looks like the client lost connection to the broker for some reason. Long
GC perhaps? But eventually the service should have recovered. Any
indication of connection
recovery later?
Jerry
On Sun, Dec 24, 2017 at 11:34 AM, Eddie Epstein wrote:
> Jerry,
>
> Testing a real
Jerry,
Testing a real analytic running 28 sync pipelines as a DUCC managed UIMA-AS
service, the DUCC pinger is reporting failures for an idle service 2 or 3
times a day.
sm.log, an error surrounded by good pings, the ping normally done every 2
minutes:
Dec 2017 07:49:42,740 INFO SM.PingDriver
- Built from source - OK
- Spot checked signatures - OK
- issues fixed - OK
- Checked README, RELEASE_NOTES,LICENSE and NOTICE - OK
- Documentation (ducc book, javadoc api) - OK
- Installed eclipse plugins in eclipse Mars 2 - OK
[ ] +1 OK to release
Jerry
On Thu, Dec 14, 2017 at 4:52 PM,
Hi,
the UIMA-AS 2.10.2 release candidate 5 is ready for voting.
This version contains the following fixes:
- fixed documentation
The list of changes in Jira:
http://issues.apache.org/jira/issues/?jql=fixVersion%20%3D%202.10.2AS%20AND%20project%20%3D%20UIMA
The source and binary zip/tars are
13 matches
Mail list logo