[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Monotone-devel] make check failures
From: |
Peter Simons |
Subject: |
[Monotone-devel] make check failures |
Date: |
18 May 2004 17:59:37 +0200 |
I am getting the following "make check" failures on two
machines of mine:
22: exchanging work via netsync FAILED (t_netsync.at:33)
26: single manifest netsync FAILED (t_singlenetsync.at:27)
I had to abort the running monotone process manually after a
couple of hours, it seemed to be caught in an endless loop
and eat up all CPU cycles.
This problem curiously enough occurs only on _old_ machines:
one of them is a Pentium 120, the other one is something
only slightly faster. And machines with faster CPUs (and an
identical installation) the self-test runs fine.
Could this be a timing problem in the netsync protocol? Or a
race condition?
The testsuite.log file says:
connecting to 127.0.0.1:5555
db.execute("ROLLBACK")
monotone: fatal: std::runtime_error: trapped network
exception: failed to connect: Connection refused
...
It appears, the server process does a listen(2), but fails
to accept(2) the connection?
Any idea what is going wrong, anyone?
Peter
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- [Monotone-devel] make check failures,
Peter Simons <=