[innogysmarthome] Reconnect fixes (#8182) #8353

Novanic · 2020-08-27T09:11:33Z

This pull-request fixes a few reconnect problems regarding the WebSocket, see issue #8182 for more information

When an error occurred while starting the websocket, no reconnect attempt was started which caused that the binding loses the connection (till the next restart of the server or binding), see issue [innogysmarthome] Insufficient reconnect attempts #8182 for more information
When an error occurs, it is now waited at least for 30 seconds for the next reconnect attempt to avoid too many requests (on re-occurring errors).
Fix for a side-effect of "Fix possible resource leak [innogysmarthome] Fix possible resource leak #8080" (the websocket were closed unnormally by default which could trigger false and about one reconnect attempt per second).
NullPointerException fixed
An Exception could cancel the scheduler, so no reconnect attempt was executed.
An Exception could get the binding in a mode which executed unnecessary reconnects at a fixed rate.
First automated tests added for the binding.

Update

…8324)" This reverts commit ce5d4d8. Signed-off-by: Sven Strohschein <[email protected]>

…ore it wasn't rescheduled, because startClient is called by a scheduler which is everytime still running, so the detection was broken) Signed-off-by: Sven Strohschein <[email protected]>

Signed-off-by: Sven Strohschein <[email protected]>

… when an error occurs it is scheduled automatically again (so it is useless and disturbing the it is re-scheduled automatically with a fixed rate). Signed-off-by: Sven Strohschein <[email protected]>

Signed-off-by: Sven Strohschein <[email protected]>

…Exception(...), the scheduler was canceled again, because the code ran normally again after the Exception handling which reaches cancelReinitJob()... That caused that the scheduler wasn't executed in this case... Signed-off-by: Sven Strohschein <[email protected]>

TravisBuddy · 2020-08-27T09:17:08Z

Travis tests were successful

Hey @Novanic,
we found no major flaws with your code. Still you might want to look at this logfile, as we usually suggest some optional improvements.

Hilbrand

LGTM. I'll keep it open for a few more days so you can test if it's running ok this time 😉

Novanic · 2020-08-29T12:29:45Z

Ok. :-)

Here is a built version with the changes for testing, some people of a OpenHAB Innogy Facebook group seem also to test it now: https://www.dropbox.com/s/5vv5qpnznelq10d/org.openhab.binding.innogysmarthome-2.5.9-SNAPSHOT.jar?dl=0

Hilbrand · 2020-09-01T08:12:14Z

Looks like you need to fix the conflicts cause by your other pr: #8375

TravisBuddy · 2020-09-02T18:13:14Z

Travis tests were successful

Hey @Novanic,
we found no major flaws with your code. Still you might want to look at this logfile, as we usually suggest some optional improvements.

Novanic · 2020-09-02T18:25:57Z

I hate Git... Just forgot to add Signed-off-by to the commit message.............. I think I have to create a new branch and a new pull-request........

Signed-off-by: Sven Strohschein <[email protected]>

TravisBuddy · 2020-09-02T18:31:59Z

Travis tests were successful

Hey @Novanic,
we found no major flaws with your code. Still you might want to look at this logfile, as we usually suggest some optional improvements.

Signed-off-by: Sven Strohschein <[email protected]>

Hilbrand · 2020-09-02T18:37:10Z

Looks like everything is ok? No need to create a new pr. When having to fix previous commits you can always use git rebase -i HEAD~<number of commits to fix>. and than push force the changes back.

TravisBuddy · 2020-09-02T18:43:36Z

Travis tests were successful

Hey @Novanic,
we found no major flaws with your code. Still you might want to look at this logfile, as we usually suggest some optional improvements.

Novanic · 2020-09-02T18:44:18Z

Yes, I had luck that there was only the last commit message without sign-off. I had problems in the past with rebase and merge (rebases and merge commits weren't recognized correctly, all commits of other people were added to the pull-request). I think the problem is that I'm working with a fork and creating pull-requests from a branch of the fork... I should give up the fork for the next pull-requests.

Novanic · 2020-09-02T18:47:19Z

I think the pull-request could now get merged. At least 3 people tested the version for a few days and there were some reconnect attempts which were all successfully. The greatest risk is that I did something wrong on fixing the branch today... ;-)

* startClient can now also get rescheduled when the method crashes (before it wasn't rescheduled, because startClient is called by a scheduler which is everytime still running, so the detection was broken) * The re-init job isn't scheduled for more than 1 time anymore, because when an error occurs it is scheduled automatically again (so it is useless and disturbing the it is re-scheduled automatically with a fixed rate). * Tests for InnogyBridgeHandler added * Bug-Fix: When an Exception occurred which was handled by handleClientException(...), the scheduler was canceled again, because the code ran normally again after the Exception handling which reaches cancelReinitJob()... That caused that the scheduler wasn't executed in this case... * Warning message improved Closes openhab#8182 Signed-off-by: Sven Strohschein <[email protected]>

Novanic and others added 30 commits June 10, 2020 19:50

Merge pull request #1 from openhab/2.5.x

78699f9

Update

Merge pull request #2 from openhab/2.5.x

ae6b891

Update

Merge pull request #3 from openhab/2.5.x

66abc11

Update

Merge pull request #4 from openhab/2.5.x

4f2cd51

Update

Merge pull request #5 from openhab/2.5.x

bd9cfc1

Update

Merge pull request #6 from openhab/2.5.x

de62002

Update

Merge pull request #7 from openhab/2.5.x

ad272b7

Update

Merge pull request #8 from openhab/2.5.x

76c47b8

Update

Merge pull request #9 from openhab/2.5.x

2ce668b

Update

Merge pull request #10 from openhab/2.5.x

bcfbee2

Update

Merge pull request #11 from openhab/2.5.x

2d654b9

Update

Merge pull request #12 from openhab/2.5.x

05de7f6

Update

Merge pull request #13 from openhab/2.5.x

77b02c4

Update

Merge pull request #14 from openhab/2.5.x

7a6babc

Update

Merge pull request #15 from openhab/2.5.x

771dda8

Update

Merge pull request #16 from openhab/2.5.x

626eeb1

Update

Revert "Revert "[innogysmarthome] Reconnect fixes (#8182) (#8318)" (#…

a58e461

…8324)" This reverts commit ce5d4d8. Signed-off-by: Sven Strohschein <[email protected]>

startClient can now also get rescheduled when the method crashes (bef…

55949a2

…ore it wasn't rescheduled, because startClient is called by a scheduler which is everytime still running, so the detection was broken) Signed-off-by: Sven Strohschein <[email protected]>

Code optimization

ed5ea82

Signed-off-by: Sven Strohschein <[email protected]>

Code optimization

0cf2b9d

Signed-off-by: Sven Strohschein <[email protected]>

The re-init job isn't scheduled for more than 1 time anymore, because…

8462cbf

… when an error occurs it is scheduled automatically again (so it is useless and disturbing the it is re-scheduled automatically with a fixed rate). Signed-off-by: Sven Strohschein <[email protected]>

Tests for InnogyBridgeHandler added

771f4c0

Signed-off-by: Sven Strohschein <[email protected]>

Tests for InnogyBridgeHandler added

bab3ff5

Signed-off-by: Sven Strohschein <[email protected]>

Code optimization

8435ff2

Signed-off-by: Sven Strohschein <[email protected]>

Code optimization

862a663

Signed-off-by: Sven Strohschein <[email protected]>

Code optimization

c3abfc5

Signed-off-by: Sven Strohschein <[email protected]>

Code optimization

d4efd37

Signed-off-by: Sven Strohschein <[email protected]>

Code optimization

6084d45

Signed-off-by: Sven Strohschein <[email protected]>

Code optimization

5ab9dab

Signed-off-by: Sven Strohschein <[email protected]>

Tests added

0408859

Signed-off-by: Sven Strohschein <[email protected]>

sstrohschein added 4 commits August 24, 2020 23:36

Tests added

a488fc1

Signed-off-by: Sven Strohschein <[email protected]>

Tests added

c607d5b

Signed-off-by: Sven Strohschein <[email protected]>

Spotless applied

1c0a07b

Signed-off-by: Sven Strohschein <[email protected]>

Tests added

1d8e2ad

Signed-off-by: Sven Strohschein <[email protected]>

Novanic requested a review from ollie-dev as a code owner August 27, 2020 09:11

Hilbrand added the bug An unexpected problem or unintended behavior of an add-on label Aug 27, 2020

openhab deleted a comment from TravisBuddy Aug 29, 2020

Hilbrand approved these changes Aug 29, 2020

View reviewed changes

Hilbrand added this to the 2.5.9 milestone Aug 29, 2020

Merge branch '2.5.x' into 8182-innogysmarthome-reconnect-3

47328ad

Update InnogyBridgeHandlerTest.java

8a28410

Signed-off-by: Sven Strohschein <[email protected]>

Warning message improved

9316b52

Signed-off-by: Sven Strohschein <[email protected]>

Hilbrand merged commit aa32b69 into openhab:2.5.x Sep 2, 2020

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[innogysmarthome] Reconnect fixes (#8182) #8353

[innogysmarthome] Reconnect fixes (#8182) #8353

Novanic commented Aug 27, 2020 •

edited

Loading

TravisBuddy commented Aug 27, 2020

Hilbrand left a comment

Novanic commented Aug 29, 2020 •

edited

Loading

Hilbrand commented Sep 1, 2020

TravisBuddy commented Sep 2, 2020

Novanic commented Sep 2, 2020

TravisBuddy commented Sep 2, 2020

Hilbrand commented Sep 2, 2020

TravisBuddy commented Sep 2, 2020

Novanic commented Sep 2, 2020

Novanic commented Sep 2, 2020

[innogysmarthome] Reconnect fixes (#8182) #8353

[innogysmarthome] Reconnect fixes (#8182) #8353

Conversation

Novanic commented Aug 27, 2020 • edited Loading

TravisBuddy commented Aug 27, 2020

Travis tests were successful

Hilbrand left a comment

Choose a reason for hiding this comment

Novanic commented Aug 29, 2020 • edited Loading

Hilbrand commented Sep 1, 2020

TravisBuddy commented Sep 2, 2020

Travis tests were successful

Novanic commented Sep 2, 2020

TravisBuddy commented Sep 2, 2020

Travis tests were successful

Hilbrand commented Sep 2, 2020

TravisBuddy commented Sep 2, 2020

Travis tests were successful

Novanic commented Sep 2, 2020

Novanic commented Sep 2, 2020

Novanic commented Aug 27, 2020 •

edited

Loading

Novanic commented Aug 29, 2020 •

edited

Loading