Skip Menu |

This queue is for tickets about the Data-Sah-Coerce CPAN distribution.

Report information
The Basics
Id: 114614
Status: resolved
Priority: 0/
Queue: Data-Sah-Coerce

People
Owner: Nobody in particular
Requestors: SREZIC [...] cpan.org
Cc:
AdminCc:

Bug Information
Severity: (no value)
Broken in: 0.011
Fixed in: (no value)



Subject: Indexer problems?
0.011 does not seem to be indexed (some eight hours or so after upload). MetaCPAN still links to the 0.010 version: https://metacpan.org/release/Data-Sah-Coerce And the CPAN.pm shell also lists the old version: cpan[2]> m Data::Sah::Coerce Module id = Data::Sah::Coerce CPAN_USERID PERLANCAR (perlancar <perlancar@gmail.com>) CPAN_VERSION 0.010 CPAN_FILE P/PE/PERLANCAR/Data-Sah-Coerce-0.010.tar.gz INST_FILE (not installed)
On Mon, 23 May 2016 18:13:34 GMT, SREZIC wrote: Show quoted text
> 0.011 does not seem to be indexed (some eight hours or so after > upload). MetaCPAN still links to the 0.010 version: > https://metacpan.org/release/Data-Sah-Coerce > And the CPAN.pm shell also lists the old version: > > cpan[2]> m Data::Sah::Coerce > Module id = Data::Sah::Coerce > CPAN_USERID PERLANCAR (perlancar <perlancar@gmail.com>) > CPAN_VERSION 0.010 > CPAN_FILE P/PE/PERLANCAR/Data-Sah-Coerce-0.010.tar.gz > INST_FILE (not installed)
PAUSE reported indexing success though. I'll reindex.
On Tue, 24 May 2016 06:08:54 GMT, PERLANCAR wrote: Show quoted text
> On Mon, 23 May 2016 18:13:34 GMT, SREZIC wrote:
> > 0.011 does not seem to be indexed (some eight hours or so after > > upload). MetaCPAN still links to the 0.010 version: > > https://metacpan.org/release/Data-Sah-Coerce > > And the CPAN.pm shell also lists the old version: > > > > cpan[2]> m Data::Sah::Coerce > > Module id = Data::Sah::Coerce > > CPAN_USERID PERLANCAR (perlancar <perlancar@gmail.com>) > > CPAN_VERSION 0.010 > > CPAN_FILE P/PE/PERLANCAR/Data-Sah-Coerce-0.010.tar.gz > > INST_FILE (not installed)
> > PAUSE reported indexing success though. I'll reindex.
Fixed, but something peculiar about why this happened. I'm cc-ing to Andreas in case he wants to investigate. I suspect in this case, 0.010 was somehow indexed *after* 0.011. Or, since that shouldn't have succeeded in the first place, perhaps at the same time (race condition)? I'm guessing the indexer runs in more than one processes/threads? Details below. Due to network flakiness on my part and also the recent PAUSE random 500 errors, I sometimes upload several distributions "at once" (within seconds from one another, in a single batch, using App::pause CLI: 'pause upload *'). And in this case, 0.010 was uploaded along with 0.011. The timestamps from PAUSE upload email report look like this: Subject: CPAN Upload: P/PE/PERLANCAR/Data-Sah-Coerce-0.011.tar.gz To: "perlancar" <perlancar@gmail.com> From: PAUSE <upload@pause.perl.org> Message-Id: <20160523062859.5E4D8195E@pause.perl.org> Date: Mon, 23 May 2016 06:28:59 +0000 (UTC) Subject: CPAN Upload: P/PE/PERLANCAR/Data-Sah-Coerce-0.010.tar.gz To: "perlancar" <perlancar@gmail.com> From: PAUSE <upload@pause.perl.org> Message-Id: <20160523062847.E6746195E@pause.perl.org> Date: Mon, 23 May 2016 06:28:47 +0000 (UTC) Which look right. Then, a couple of minutes later, came the PAUSE indexing report emails, which also look right (down to the timestamps of all the Received headers): Subject: PAUSE indexer report PERLANCAR/Data-Sah-Coerce-0.011.tar.gz From: PAUSE <upload@pause.perl.org> Date: Tue, 24 May 2016 06:56:16 +0000 Message-Id: <20160524065616.71524153F@pause.perl.org> Subject: PAUSE indexer report PERLANCAR/Data-Sah-Coerce-0.010.tar.gz From: PAUSE <upload@pause.perl.org> Date: Tue, 24 May 2016 06:55:47 +0000 Message-Id: <20160524065547.989BC153F@pause.perl.org> The only strange part is that Gmail sorts the 0.010 email after 0.011, which means it thinks the 0.010 email came after 0.011. After the reindexing, Gmail sorts the 0.011 report email after 0.010. And this time 0.011 is indexed as the latest version.