[
https://issues.jboss.org/browse/JBIDE-23422?page=com.atlassian.jira.plugi...
]
Jeff Cantrill commented on JBIDE-23422:
---------------------------------------
Maybe a larger topic exposed by this issue, but we should revisit how we debug since I
dont think the router should need to be involved. Besides the issues described above, its
an extra network hop, it places you behind the service, and doesn't guarantee a
specific pod to debug. We should be able to accomplish debug using port forwarding which
does not depend on a route or service. We should additionally be able to accomplish pod
selection or inference with minimal input from the user. Changing the router timeouts is
not going to be available to the non-admin user unless they are working against a dev
instance of openshift.
Server Adapter: Node.js debug session is terminated after ~1 minute
and browser shows 502 error
-----------------------------------------------------------------------------------------------
Key: JBIDE-23422
URL:
https://issues.jboss.org/browse/JBIDE-23422
Project: Tools (JBoss Tools)
Issue Type: Bug
Components: javascript, openshift
Affects Versions: 4.4.2.AM2
Environment: windows 10
Fedora 24
Reporter: Ilya Buziuk
Assignee: Ilya Buziuk
Priority: Critical
Labels: nodejs, openshift_v3, server_adapter
Fix For: 4.4.2.Final
We've found, that when one is debugging (the code is stopeed at breakpoint, stepping
through code, inspecting variables, ...) longer than ~1 minute, browser displays error
code 502 and the debug session gets terminated. This makes this feature not very useful,
because all debugging must be quicker than that timeout.
The behavior is captured in this screencast:
https://youtu.be/BJf7wcPqNmM (note how at
the time 0:42 the page is finally loaded (502 error) and the debug session is terminated
(in debug view)).
We were able to reproduce this issue on F24 and Win10 (using CDK and
console.engint.openshift.com)
--
This message was sent by Atlassian JIRA
(v7.2.3#72005)