[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#929029: marked as done ([pre-approval] unblock: apt-cacher-ng/3.2.1-1)



Your message dated Sat, 25 May 2019 18:41:50 +0100
with message-id <20190525174150.GA8108@powdarrmonkey.net>
and subject line Re: Bug#929029: unblock: apt-cacher-ng/3.2.1-1
has caused the Debian Bug report #929029,
regarding [pre-approval] unblock: apt-cacher-ng/3.2.1-1
to be marked as done.

This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.

(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact owner@bugs.debian.org
immediately.)


-- 
929029: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=929029
Debian Bug Tracking System
Contact owner@bugs.debian.org with problems
--- Begin Message ---
Package: release.debian.org
Severity: normal
User: release.debian.org@packages.debian.org
Usertags: unblock

Please advise how to process with a required fix for the package apt-cacher-ng.

The change is a one-liner and solves the bug #928957. Without it, the
cache cleanup will fail for a lot of people in the next couple of years.
But it touches the upstream source, that's why I would like to release
it as minor upstream version (3.2.1, currently 3.2, and I am the
upstream).

I remember how you handled a similar request of mine a couple of years
ago, and this time I DEMAND a proper response here before I upload
anything. Please don't ignore it again for weeks and don't tell me that
this change is impossible to understand or to estimate WRT consequences;
it is using an already existing interface in the exact usecase it was
designed for. (see below)

Best regards,
Eduard.

diff --git a/ChangeLog b/ChangeLog
index c064c44..67e440a 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,10 +1,20 @@
+apt-cacher-ng (3.2.1) MY-NAME-IS-ANYBODY-3D; urgency=low
+
+  * Fix overoptimistic guess of the SHA256SUMS file location of Debian
+    repositories, which made the expiration task fail without any way to
+    recover from it. Now ignoring a download error in this case since
+    permanent 404ing for other reasons than loss of remote content (on any
+    mirror) can be considered unlikely.
+
+ -- Eduard Bloch <blade@debian.org>  Wed, 15 May 2019 20:32:44 +0200
+
 apt-cacher-ng (3.2) MY-NAME-IS-ANYBODY; urgency=medium

   * Maintenance release
   * Basic support for FlatPak repositories
   * Added flat by-hash structure to file patterns
   * Allowing some Debian and Ubuntu services in default PassThroughPattern
   * Various fixes in manpage and configuration examples
   * Mirror database update

  -- Eduard Bloch <blade@debian.org>  Fri, 07 Sep 2018 12:55:46 +0200
diff --git a/source/expiration.cc b/source/expiration.cc
index 2125521..625a7be 100644
--- a/source/expiration.cc
+++ b/source/expiration.cc
@@ -799,20 +799,27 @@ bool expiration::ProcessRegular(const string & sPathAbs, const struct stat &stin
 		auto& flags = m_metaFilesRel[idir +"SHA256SUMS"];

 		/* pretend that it's there but not usable so the refreshing code will try to get at
 		 * least one copy for that location if it's needed there
 		 */
 		if(!flags.vfile_ondisk)
 		{
 			flags.eIdxType = EIDX_SHA256DILIST;
 			flags.vfile_ondisk = true;
 			flags.uptodate = false;
+
+			// the original source context will probably provide a viable source for
+			// this URL - it might go 404 if the whole folder is missing but then the
+			// referenced content would also be outdated/gone and not worth keeping
+			// in the cache anyway
+
+			flags.forgiveDlErrors = true;
 		}
 		// and last but not least - care only about the modern version of that index
 		m_metaFilesRel.erase(idir + "MD5SUMS");
 	}
 	unsigned stripLen=0;
     if (endsWithSzAr(sPathRel, ".head"))
 		stripLen=5;
 	else if (AddIFileCandidate(sPathRel))
 	{
 		auto &attr = SetFlags(sPathRel);

--- End Message ---
--- Begin Message ---
Control: tag -1 - moreinfo

Hi,

On Sat, May 18, 2019 at 08:54:00AM +0000, Niels Thykier wrote:
> Please go ahead with the change and remove the moreinfo tag when it has
> been uploaded to unstable and it is ready to be unblocked.

Although moreinfo is still present this seems to have been uploaded and I
have unblocked it.

Thanks,

-- 
Jonathan Wiltshire                                      jmw@debian.org
Debian Developer                         http://people.debian.org/~jmw

4096R: 0xD3524C51 / 0A55 B7C5 1223 3942 86EC  74C3 5394 479D D352 4C51

--- End Message ---

Reply to: