Tony Tkacik -X (ttkacik - Pantheon Technologies SRO@Cisco) <ttkacik@...>
Hi,
I am interested to run / execute roboto test locally,
To reproduce errors you are seeing in roboto
Tony
|
|
Tony Tkacik -X (ttkacik - Pantheon Technologies SRO@Cisco) <ttkacik@...>
Hi,
The reason for this is instability in openflow plugin test-cases,
I noticed several run of same controller build (triggered by ovsdb or bgpcep)
yelded different results, this may be caused since MD-SAL and Openflow 1.3
is highly asynchronous, I noticed that tests sometimes waits and try, which
may sometimes not succeed.
So I want to identify this potential tests and figure out how to improve
testing of asynchronous actions.
Tony
toggle quoted message
Show quoted text
From: Tony Tkacik
-X (ttkacik - Pantheon Technologies SRO at Cisco)
Sent: Wednesday, May 07, 2014 6:40 PM
To: integration-dev@...
Subject: Running Roboto Tests locally?
Hi,
I am interested to run / execute roboto test locally,
To reproduce errors you are seeing in roboto
Tony
|
|
Christopher O'SHEA <christopher.o.shea@...>
Hi Tony,
pybot -v CONTROLLER:127.0.0.1 -v MININET:127.0.0.1 -v USER_HOME:${HOME} -v MININET_USER:${USER} ${HOME}/integration/test/csit/suites/base/
Should be the command to run,
Just need to make sure your SSH key exchange so that robot can ssh into localhost to start mini net.
Chris
Hi,
The reason for this is instability in openflow plugin test-cases,
I noticed several run of same controller build (triggered by ovsdb or bgpcep)
yelded different results, this may be caused since MD-SAL and Openflow 1.3
is highly asynchronous, I noticed that tests sometimes waits and try, which
may sometimes not succeed.
So I want to identify this potential tests and figure out how to improve
testing of asynchronous actions.
Tony
toggle quoted message
Show quoted text
From: Tony Tkacik -X (ttkacik - Pantheon Technologies SRO at Cisco)
Sent: Wednesday, May 07, 2014 6:40 PM
To: integration-dev@...
Subject: Running Roboto Tests locally?
Hi,
I am interested to run / execute roboto test locally,
To reproduce errors you are seeing in roboto
Tony
|
|
Hi Tony,
I agree most of the problems we have now are related to MD-SAL async behavior and I am also trying to identify if there is any other issue that makes the test suite so unstable.
I can have a dedicated chat with you to explain you how to setup the system test in your local. You can start by downloading and reading this wiki and also watching at the TWS recording at the bottom:
BR/Luis On May 7, 2014, at 9:59 AM, Tony Tkacik -X (ttkacik - Pantheon Technologies SRO at Cisco) < ttkacik@...> wrote:
toggle quoted message
Show quoted text
Hi, The reason for this is instability in openflow plugin test-cases, I noticed several run of same controller build (triggered by ovsdb or bgpcep) yelded different results, this may be caused since MD-SAL and Openflow 1.3 is highly asynchronous, I noticed that tests sometimes waits and try, which may sometimes not succeed. So I want to identify this potential tests and figure out how to improve testing of asynchronous actions. Tony From: Tony Tkacik -X (ttkacik - Pantheon Technologies SRO at Cisco) Sent: Wednesday, May 07, 2014 6:40 PM To: integration-dev@... Subject: Running Roboto Tests locally? Hi, I am interested to run / execute roboto test locally, To reproduce errors you are seeing in roboto Tony _______________________________________________ integration-dev mailing list integration-dev@...https://lists.opendaylight.org/mailman/listinfo/integration-dev
|
|
Hi Tony, I have another question for you: are most all the OSGi errors and exceptions we see with OF13 also due to MD-SAL async behavior?
BR/Luis
toggle quoted message
Show quoted text
On May 7, 2014, at 10:27 AM, Luis Gomez < ecelgp@...> wrote: Hi Tony,
I agree most of the problems we have now are related to MD-SAL async behavior and I am also trying to identify if there is any other issue that makes the test suite so unstable.
I can have a dedicated chat with you to explain you how to setup the system test in your local. You can start by downloading and reading this wiki and also watching at the TWS recording at the bottom:
BR/Luis On May 7, 2014, at 9:59 AM, Tony Tkacik -X (ttkacik - Pantheon Technologies SRO at Cisco) < ttkacik@...> wrote: Hi, The reason for this is instability in openflow plugin test-cases, I noticed several run of same controller build (triggered by ovsdb or bgpcep) yelded different results, this may be caused since MD-SAL and Openflow 1.3 is highly asynchronous, I noticed that tests sometimes waits and try, which may sometimes not succeed. So I want to identify this potential tests and figure out how to improve testing of asynchronous actions. Tony From: Tony Tkacik -X (ttkacik - Pantheon Technologies SRO at Cisco) Sent: Wednesday, May 07, 2014 6:40 PM To: integration-dev@... Subject: Running Roboto Tests locally? Hi, I am interested to run / execute roboto test locally, To reproduce errors you are seeing in roboto Tony _______________________________________________ integration-dev mailing list integration-dev@...https://lists.opendaylight.org/mailman/listinfo/integration-dev
|
|
Hey Luis,
not sure which errors/exceptions you are referring to. Got any
pointers?
Thanks,
Robert
On 05/07/2014 08:57 PM, Luis Gomez wrote:
toggle quoted message
Show quoted text
Hi Tony, I have another question for you: are most all the
OSGi errors and exceptions we see with OF13 also due to MD-SAL
async behavior?
BR/Luis
On May 7, 2014, at 10:27 AM, Luis Gomez < ecelgp@...>
wrote:
Hi Tony,
I agree most of the problems we have now are related to
MD-SAL async behavior and I am also trying to identify if
there is any other issue that makes the test suite so
unstable.
I can have a dedicated chat with you to explain you how
to setup the system test in your local. You can start by
downloading and reading this wiki and also watching at the
TWS recording at the bottom:
BR/Luis
On May 7, 2014, at 9:59 AM, Tony Tkacik -X
(ttkacik - Pantheon Technologies SRO at Cisco) < ttkacik@...>
wrote:
Hi,
The reason
for this is instability in openflow plugin
test-cases,
I noticed
several run of same controller build
(triggered by ovsdb or bgpcep)
yelded
different results, this may be caused since
MD-SAL and Openflow 1.3
is highly
asynchronous, I noticed that tests sometimes
waits and try, which
may
sometimes not succeed.
So I want
to identify this potential tests and figure
out how to improve
testing of
asynchronous actions.
Tony
From: Tony
Tkacik -X (ttkacik - Pantheon
Technologies SRO at Cisco)
Sent: Wednesday,
May 07, 2014 6:40 PM
To: integration-dev@...
Subject: Running
Roboto Tests locally?
Hi,
I am
interested to run / execute roboto test
locally,
To
reproduce errors you are seeing in roboto
Tony
_______________________________________________
integration-dev mailing list
integration-dev@...
https://lists.opendaylight.org/mailman/listinfo/integration-dev
_______________________________________________
integration-dev mailing list
integration-dev@...
https://lists.opendaylight.org/mailman/listinfo/integration-dev
|
|
Hi Robert,
This link shows real-time OSGi console in our test automation:
BR/Luis
toggle quoted message
Show quoted text
On May 7, 2014, at 1:18 PM, Robert Varga < nite@...> wrote:
Hey Luis,
not sure which errors/exceptions you are referring to. Got any
pointers?
Thanks,
Robert
On 05/07/2014 08:57 PM, Luis Gomez wrote:
Hi Tony, I have another question for you: are most all the
OSGi errors and exceptions we see with OF13 also due to MD-SAL
async behavior?
BR/Luis
On May 7, 2014, at 10:27 AM, Luis Gomez < ecelgp@...>
wrote:
Hi Tony,
I agree most of the problems we have now are related to
MD-SAL async behavior and I am also trying to identify if
there is any other issue that makes the test suite so
unstable.
I can have a dedicated chat with you to explain you how
to setup the system test in your local. You can start by
downloading and reading this wiki and also watching at the
TWS recording at the bottom:
BR/Luis
On May 7, 2014, at 9:59 AM, Tony Tkacik -X
(ttkacik - Pantheon Technologies SRO at Cisco) < ttkacik@...>
wrote:
Hi,
The reason
for this is instability in openflow plugin
test-cases,
I noticed
several run of same controller build
(triggered by ovsdb or bgpcep)
yelded
different results, this may be caused since
MD-SAL and Openflow 1.3
is highly
asynchronous, I noticed that tests sometimes
waits and try, which
may
sometimes not succeed.
So I want
to identify this potential tests and figure
out how to improve
testing of
asynchronous actions.
Tony
From: Tony
Tkacik -X (ttkacik - Pantheon
Technologies SRO at Cisco)
Sent: Wednesday,
May 07, 2014 6:40 PM
To: integration-dev@...
Subject: Running
Roboto Tests locally?
Hi,
I am
interested to run / execute roboto test
locally,
To
reproduce errors you are seeing in roboto
Tony
_______________________________________________
integration-dev mailing list
integration-dev@...
https://lists.opendaylight.org/mailman/listinfo/integration-dev
_______________________________________________
integration-dev mailing list
integration-dev@...
https://lists.opendaylight.org/mailman/listinfo/integration-dev
|
|
I have to admit it is much better now than a few days ago…
toggle quoted message
Show quoted text
On May 7, 2014, at 1:23 PM, Luis Gomez < ecelgp@...> wrote:
Hi Robert,
This link shows real-time OSGi console in our test automation:
BR/Luis
On May 7, 2014, at 1:18 PM, Robert Varga < nite@...> wrote:
Hey Luis,
not sure which errors/exceptions you are referring to. Got any
pointers?
Thanks,
Robert
On 05/07/2014 08:57 PM, Luis Gomez wrote:
Hi Tony, I have another question for you: are most all the
OSGi errors and exceptions we see with OF13 also due to MD-SAL
async behavior?
BR/Luis
On May 7, 2014, at 10:27 AM, Luis Gomez < ecelgp@...>
wrote:
Hi Tony,
I agree most of the problems we have now are related to
MD-SAL async behavior and I am also trying to identify if
there is any other issue that makes the test suite so
unstable.
I can have a dedicated chat with you to explain you how
to setup the system test in your local. You can start by
downloading and reading this wiki and also watching at the
TWS recording at the bottom:
BR/Luis
On May 7, 2014, at 9:59 AM, Tony Tkacik -X
(ttkacik - Pantheon Technologies SRO at Cisco) < ttkacik@...>
wrote:
Hi,
The reason
for this is instability in openflow plugin
test-cases,
I noticed
several run of same controller build
(triggered by ovsdb or bgpcep)
yelded
different results, this may be caused since
MD-SAL and Openflow 1.3
is highly
asynchronous, I noticed that tests sometimes
waits and try, which
may
sometimes not succeed.
So I want
to identify this potential tests and figure
out how to improve
testing of
asynchronous actions.
Tony
From: Tony
Tkacik -X (ttkacik - Pantheon
Technologies SRO at Cisco)
Sent: Wednesday,
May 07, 2014 6:40 PM
To: integration-dev@...
Subject: Running
Roboto Tests locally?
Hi,
I am
interested to run / execute roboto test
locally,
To
reproduce errors you are seeing in roboto
Tony
_______________________________________________
integration-dev mailing list
integration-dev@...
https://lists.opendaylight.org/mailman/listinfo/integration-dev
_______________________________________________
integration-dev mailing list
integration-dev@...
https://lists.opendaylight.org/mailman/listinfo/integration-dev
|
|
Hi Luis,
thanks for the link. Quick review ...
- the initial warnings are obviously from the container startup,
not harmful
- InterruptedExceptions -- I think Michal Rehak was working around
that (if I overheard and remembered the conversation correctly)
- the did not pass canCommit/commit failed messages come from
asynchronous writes and strict datastore. they usually indicate
the data being written has been modified while the write was being
set up, hinting at unsychronized or racy access
- the session cookie exceptions seem to be normal if switches come
and go -- maybe we could downgrade them
By the looks of it this should not be hard to get stabilized as a
matter of two-three weeks -- unless they end up uncovering
something very significantly wrong in app design.
Bye,
Robert
On 05/07/2014 10:23 PM, Luis Gomez wrote:
toggle quoted message
Show quoted text
Hi Robert,
This link shows real-time OSGi console in our test
automation:
BR/Luis
On May 7, 2014, at 1:18 PM, Robert Varga < nite@...>
wrote:
Hey Luis,
not sure which errors/exceptions you are referring to. Got
any pointers?
Thanks,
Robert
On 05/07/2014 08:57 PM, Luis Gomez wrote:
Hi Tony, I have another question for you: are most
all the OSGi errors and exceptions we see with OF13 also
due to MD-SAL async behavior?
BR/Luis
On May 7, 2014, at 10:27 AM, Luis Gomez < ecelgp@...>
wrote:
Hi
Tony,
I agree most of the problems we have now are
related to MD-SAL async behavior and I am also
trying to identify if there is any other issue
that makes the test suite so unstable.
I can have a dedicated chat with you to explain
you how to setup the system test in your local.
You can start by downloading and reading this wiki
and also watching at the TWS recording at the
bottom:
BR/Luis
On May 7, 2014, at 9:59 AM, Tony Tkacik
-X (ttkacik - Pantheon Technologies SRO at
Cisco) < ttkacik@...>
wrote:
Hi,
The reason for
this is instability in openflow
plugin test-cases,
I noticed several
run of same controller build
(triggered by ovsdb or bgpcep)
yelded different
results, this may be caused since
MD-SAL and Openflow 1.3
is highly
asynchronous, I noticed that tests
sometimes waits and try, which
may sometimes not
succeed.
So I want to
identify this potential tests and
figure out how to improve
testing of
asynchronous actions.
Tony
From: Tony
Tkacik -X (ttkacik - Pantheon
Technologies SRO at Cisco)
Sent: Wednesday,
May 07, 2014 6:40 PM
To: integration-dev@...
Subject: Running
Roboto Tests locally?
Hi,
I am interested to run /
execute roboto test locally,
To reproduce errors you
are seeing in roboto
Tony
_______________________________________________
integration-dev mailing list
integration-dev@...
https://lists.opendaylight.org/mailman/listinfo/integration-dev
_______________________________________________
integration-dev mailing list
integration-dev@...
https://lists.opendaylight.org/mailman/listinfo/integration-dev
|
|