https://bz.apache.org/SpamAssassin/show_bug.cgi?id=8227
Bug ID: 8227
Summary: Detect bad DNS network setup early in build
Product: Spamassassin
Version: SVN Trunk (Latest Devel Version)
Hardware: All
OS: All
Status: NEW
Severity: normal
Priority: P2
Component: Building & Packaging
Assignee: dev@spamassassin.apache.org
Reporter: sidney@sidney.com
Target Milestone: Undefined
With the fixes for bug 8225 and 8213, and maybe bug 8204, I'm seeing test
failures on some test machines on CPAN that are running FreeBSD 10.1 and some
but not all Windows test machines. I think the failures are symptoms of a DNS
resolver not being set up correctly, and may explain the more elusive failures
I've been trying to track down in bug 8201 and bug 8198.
I'm adding a new test t/dnsplatform.t to check that the testing platform does
support large DNS reply packets as well as the multihomed query that spamc_H.t
relies on. I'm also adding a similar test to DependencyInfo.pm to be called in
Makefile.PL. That will make the build fail early if the network does not have a
proper DNS resolving setup. If Net::DNS is not installed, this test will be
skipped, allowing a workaround if someone really wants to build and install
SpamAssassin on a system without proper support for the DNS functionality,
e.g., if they intend to run it only with -L, or install first and test network
settings later. By failing the build that early, CPAN test machines that don't
have proper DNS will not get far enough to generate all those reports of
failing t/askdns.t and so forth that I've been trying to solve.
--
You are receiving this mail because:
You are the assignee for the bug.
Bug ID: 8227
Summary: Detect bad DNS network setup early in build
Product: Spamassassin
Version: SVN Trunk (Latest Devel Version)
Hardware: All
OS: All
Status: NEW
Severity: normal
Priority: P2
Component: Building & Packaging
Assignee: dev@spamassassin.apache.org
Reporter: sidney@sidney.com
Target Milestone: Undefined
With the fixes for bug 8225 and 8213, and maybe bug 8204, I'm seeing test
failures on some test machines on CPAN that are running FreeBSD 10.1 and some
but not all Windows test machines. I think the failures are symptoms of a DNS
resolver not being set up correctly, and may explain the more elusive failures
I've been trying to track down in bug 8201 and bug 8198.
I'm adding a new test t/dnsplatform.t to check that the testing platform does
support large DNS reply packets as well as the multihomed query that spamc_H.t
relies on. I'm also adding a similar test to DependencyInfo.pm to be called in
Makefile.PL. That will make the build fail early if the network does not have a
proper DNS resolving setup. If Net::DNS is not installed, this test will be
skipped, allowing a workaround if someone really wants to build and install
SpamAssassin on a system without proper support for the DNS functionality,
e.g., if they intend to run it only with -L, or install first and test network
settings later. By failing the build that early, CPAN test machines that don't
have proper DNS will not get far enough to generate all those reports of
failing t/askdns.t and so forth that I've been trying to solve.
--
You are receiving this mail because:
You are the assignee for the bug.