gzserver process does not always shut down cleanly about bwi-update

[jack-oquin]

I managed to reproduce this problem running the ROS Gazebo tutorial, without any BWI code. So, it looks like a gazebo bug.

I want to leave this issue open a while, until I report it upstream.

[jack-oquin]

I think it may just take a long time for gzserver to shut down, causing the timeout.

If the user get impatient and sends a terminate signal to roslaunch, the gzserver process seems not to get cleaned up. From then on, gazebo will not run until that process stops.

[jack-oquin]

There is already an open gazebo ticket for this: https://bitbucket.org/osrf/gazebo/issue/441/frequent-escalate-to-sigkill-on-shutdown

[piyushk]

I am going to close this issue here. I can't remember why we have this issue in the bwi-update tool issues page.

[jack-oquin]

It was to track the gazebo ticket, linked above.

If we are not concerned about when it gets fixed, there is no reason to leave this one open.

[piyushk]

I can't seem to reproduce the issue. It only comes up sporadically.

On Thu, May 16, 2013 at 12:10 PM, jack-oquin [email protected]:

It was to track the gazebo ticket, linked above.

If we are not concerned about when it gets fixed, there is no reason to
leave this one open.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/6#issuecomment-18014886
.

[jack-oquin]

Let's not worry about it. The Gazebo developers will probably fix it eventually.