Doctested all of Sage, four doctests timed out—anything to be concerned about?

[Christopher Phoenix]

I just ran ./sage --testall to test my new build of Sage like the installation guide recommended. It took over six hours on my hardware (Lenovo Thinkpad 11e, Intel Celeron N2940, 4 gb ram, 500 gb hdd). When the tests finished, four doctests were listed as having timed out.

sage -t local/lib/python2.7/site-packages/sagenb/testing/tests/test_worksheet_list.py
    [0 tests, 0.19 s]
----------------------------------------------------------------------
sage -t src/sage/manifolds/differentiable/tensorfield.py  # Timed out
sage -t src/sage/manifolds/differentiable/metric.py  # Timed out
sage -t src/sage/manifolds/differentiable/affine_connection.py  # Timed out
sage -t src/sage/misc/temporary_file.py  # Timed out after testing finished
----------------------------------------------------------------------
Total time for all tests: 22027.6 seconds
    cpu time: 17542.6 seconds
    cumulative wall time: 20277.6 seconds

I guess that's not to many considering the size of the Sage library, but is it anything to be concerned about? Should I run the tests again on them? Also, I'm assuming Sage only lists tests that failed or timed out. The output doesn't say anything about the tests that succeeded, but it ran through the rest of them without complaint. Is this the case? Since Sage lists these as having "timed out", I'm guessing that testing timed out on them after they took too long, but didn't get to run through to a failure.

Partway through testing, Ubuntu reported that my system's gdb had unexpectedly quit, so perhaps that had something to do with one or some of these time-outs.

[Dima Pasechnik]

Indeed, only failed tests are listed in this summary at the end. You can see the full tests logs in logs/ subdirectory. 

You might re-run these (timed out) tests separately. However, it's not uncommon to see these, in particular if you run tests in parallel on a loaded machine. 

[Christopher Phoenix]

Thanks for the answer! :) I'll try re-running the timed-out tests separately later (tonight, maybe). Right now I'm busy working through the tuts.
I'm not sure if the tests were running in parallel or not (I just ran ./sage --testall in the sage directory), but it was pretty busy for quite some hours. I'll see what happens when I run the tests separately.

[Dima Pasechnik]

On Saturday, July 15, 2017 at 2:00:20 AM UTC+1, Christopher Phoenix wrote:

Thanks for the answer! :) I'll try re-running the timed-out tests separately later (tonight, maybe). Right now I'm busy working through the tuts.
I'm not sure if the tests were running in parallel or not (I just ran ./sage --testall in the sage directory), but it was pretty busy for quite some hours. I'll see what happens when I run the tests separately.

you can use a command like "top" in a separate shell to see what's going on on the machine, how many

CPUs are occupied and with what.