Skip Menu |

This queue is for tickets about the Net-SIP CPAN distribution.

Report information
The Basics
Id: 125624
Status: open
Priority: 0/
Queue: Net-SIP
People
Owner: Nobody in particular
Requestors: SREZIC [...] cpan.org
Cc:
AdminCc:
Bug Information
Severity: (no value)
Broken in: 0.815
Fixed in: (no value)

History Show all quoted text
  Tue Jun 19 01:38:35 2018 SREZIC [...] cpan.org - Ticket created
Subject: Random test failures
The test suite randomly fails on my smoker systems. An overview of CPAN Testers reports with sample fail reports: http://matrix.cpantesters.org/?dist=Net-SIP%200.815
  Wed Aug 15 12:34:22 2018 Steffen_Ullrich [...] genua.de - Correspondence added
On Tue Jun 19 01:38:35 2018, SREZIC wrote: Show quoted text
> The test suite randomly fails on my smoker systems. An overview of > CPAN Testers reports with sample fail reports: > http://matrix.cpantesters.org/?dist=Net-SIP%200.815
It is hard to figure out from the reports what's going on but most of the failure reports come from you and there are many successful reports. Given that some tests have a sensitive timing and the tests use multiple processes I could only imagine for now that the load on your test system is high enough that process scheduling is not as fast as I would expect and thus events do not happen within the expected time. Does this make sense in your opinion?
  Wed Aug 15 12:34:22 2018 The RT System itself - Status changed from 'new' to 'open'
  Thu Aug 16 12:58:18 2018 SREZIC [...] cpan.org - Correspondence added
On 2018-08-15 12:34:22, SULLR wrote: Show quoted text
> On Tue Jun 19 01:38:35 2018, SREZIC wrote:
> > The test suite randomly fails on my smoker systems. An overview of > > CPAN Testers reports with sample fail reports: > > http://matrix.cpantesters.org/?dist=Net-SIP%200.815
> > > It is hard to figure out from the reports what's going on but most of > the failure reports come from you and there are many successful > reports. Given that some tests have a sensitive timing and the tests > use multiple processes I could only imagine for now that the load on > your test system is high enough that process scheduling is not as fast > as I would expect and thus events do not happen within the expected > time. > Does this make sense in your opinion?
Most fail reports come from the VM identifiable as "uname=linux eserte" (per "perl -V"), which is a 1 CPU machine with 32 GB RAM, and from time to time this machine has a load average above 1. But there are also rare failures on larger machines, e.g. "linux centos6" and "linux debian9" (both with 4 CPUs and 4GB RAM) send both a fail report. On the other hand, none of my freebsd VMs send a fail report, and these are rather small (1 CPU, 1-2 GB RAM).