It could be Cinnamon crashing, or the X server. Logs are in
/tmp/chrome_remote_desktop_*
(pick the file that corresponds to the time of the issue)

I think CRD might stop logging the session after 30s, so restart CRD and try to repro the problem:
https://cs.chromium.org/search/?q=SESSION_OUTPUT_TIME_LIMIT_SECONDS&sq=package:chromium&type=cs

Filed  issue 749215  for the 30s cutoff problem.

The script is installed at /opt/google/chrome-remote-desktop/chrome-remote-desktop
just in case you need to manually edit it and change the value of SESSION_OUTPUT_TIME_LIMIT_SECONDS.

When the problem occurs I get the following written to /tmp/chrome_remote_desktop_20170726_124526_ngTJE5:

Backtrace:
0: /usr/bin/Xvfb-randr (xorg_backtrace+0x26) [0x5558b3ec8496]
1: /usr/bin/Xvfb-randr (0x5558b3d60000+0x16c33a) [0x5558b3ecc33a]
2: /lib/x86_64-linux-gnu/libpthread.so.0 (0x7f717a16b000+0x10330) [0x7f717a17b330]
3: /usr/bin/Xvfb-randr (0x5558b3d60000+0x4a088) [0x5558b3daa088]
4: /usr/bin/Xvfb-randr (0x5558b3d60000+0x113af1) [0x5558b3e73af1]
5: /usr/bin/Xvfb-randr (0x5558b3d60000+0x2e45a) [0x5558b3d8e45a]
6: /lib/x86_64-linux-gnu/libc.so.6 (__libc_start_main+0xf5) [0x7f7178fd2f45]
7: /usr/bin/Xvfb-randr (0x5558b3d60000+0x2e74d) [0x5558b3d8e74d]
Segmentation fault at address 0x14

Caught signal 11 (Segmentation fault). Server aborting

(chrome-remote-desktop-host:32268): Gdk-WARNING **: chrome-remote-desktop-host: Fatal IO error 11 (Resource temporarily unavailable) on X server :20.

2017-07-26 12:46:27,323:INFO:wait() returned (32239,256)
2017-07-26 12:46:27,323:INFO:X server process terminated
2017-07-26 12:46:27,323:INFO:Failure count for 'X server' is now 0
2017-07-26 12:46:27,323:INFO:Terminating X session
2017-07-26 12:46:27,324:INFO:wait() returned (32266,0)
2017-07-26 12:46:27,324:INFO:Session process terminated
2017-07-26 12:46:27,324:INFO:Relaunching self
2017-07-26 12:46:27,324:INFO:Cleanup.
2017-07-26 12:46:27,324:INFO:Terminating host
2017-07-26 12:46:27,436:INFO:Daemon process started in the background, logging to '/tmp/chrome_remote_desktop_20170726_124526_ngTJE5'
2017-07-26 12:46:27,437:INFO:Using host_id: 9ec44664-8ed2-c9cd-374c-a8f43514722f
...

Thanks for the log. Xvfb-randr is the component that is crashing.
It is a Google-internal patched version of Xvfb that our team built years ago.
More recent versions of Xvfb (such as in Ubuntu 16.04) have the patch already built in.

You could try removing the xvfb-randr package, which will cause plain Xvfb to be used instead. We don't have any way to symbolize or debug the backtrace for Xvfb-randr. But if you can repro the same issue with stock Xvfb, that would open the door to using a debug symbolized version.

We have a long-standing bug for Xvfb-randr crashes at b/26047696 (and b/27309469). We've never been able to get much traction on it, as the crash is so rare and difficult to repro. It's possible that yours could be the same problem.

I removed the xvfb-randr package and now it doesn't crash.

However I notice that when I finish dragging the window and let go of the mouse button, the pointer doesn't change back from the hand to the arrow the way it's supposed to.  This is the same moment (releasing mouse button) when the crash occurs with xvfb-randr.

So, I suspect it has something to do with the way Chrome changes the mouse pointer icon.

Using plain Xvfb isn't a good solution for me, since the desktop doesn't resize properly.

That's interesting to know. The crashes we've seen (where we've been able to symbolize the backtrace) occurred around the cursor-setting code (more details in the internal bug).
It looks like the Xorg implementation of setting the cursor is flaky. As it's a bug in X, it's not something we can easily address in CRD.
Best we can do is to try to repro the problem, identify the buggy code, and file an upstream X bug (possibly with a patch). But we haven't had much success so far.

Do you only see the bug when dragging Chrome windows, or with other apps?

It only happens with Chrome windows.  This makes me think it could be fixed in Chrome (by using X in a way that matches other apps).

I bisected this to http://crrev.com/a7a78c24d346 (enabling linux_aura).

Cc'ing some folks who know about X11 stuff.

Adding current linux maintainer.

I don't think there's anything we can do about this within Chromoting. Removing that component.

Updating the client system from Trusty to Rodete seemed to fix this.

Un-cc-ing me from all bugs on my final day.

Archiving old bugs that haven't been actively assigned in over 180 days.

If you feel this issue should still be addressed, feel free to reopen it or to file a new issue. Thanks!

Archiving old bugs that haven't been actively assigned in over 180 days.

If you feel this issue should still be addressed, feel free to reopen it or to file a new issue. Thanks!