Bug 231025 - devel/arduino-core failure during fetch phase during poudriere build
Summary: devel/arduino-core failure during fetch phase during poudriere build
Status: New
Alias: None
Product: Ports & Packages
Classification: Unclassified
Component: Individual Port(s) (show other bugs)
Version: Latest
Hardware: amd64 Any
: --- Affects Many People
Assignee: Kyle Evans
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2018-08-30 15:49 UTC by freebsd
Modified: 2018-09-12 02:40 UTC (History)
1 user (show)

See Also:
bugzilla: maintainer-feedback? (kevans)


Attachments
svn(1) diff against the ports tree (1.22 KB, patch)
2018-09-12 02:40 UTC, Kyle Evans
kevans: maintainer-approval+
Details | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description freebsd 2018-08-30 15:49:21 UTC
The fetch phase is failing due to file size mismatches.  Assuming the hashes will also be wrong if it got that far.


=======================<phase: fetch          >============================
===>  License LGPL21 accepted by the user
=> arduino-Arduino-1.6.18-b439a77_GH0.tar.gz doesn't seem to exist in /portdistfiles/arduino.
=> Attempting to fetch https://codeload.github.com/arduino/Arduino/tar.gz/b439a77?dummy=/arduino-Arduino-1.6.18-b439a77_GH0.tar.gz
fetch: https://codeload.github.com/arduino/Arduino/tar.gz/b439a77?dummy=/arduino-Arduino-1.6.18-b439a77_GH0.tar.gz: size unknown
fetch: https://codeload.github.com/arduino/Arduino/tar.gz/b439a77?dummy=/arduino-Arduino-1.6.18-b439a77_GH0.tar.gz: size of remote file is not known
arduino-Arduino-1.6.18-b439a77_GH0.tar.gz                0  B    0  Bps
=> Fetched file size mismatch (expected 48187661, actual 48187647)
=> Trying next site
=> Attempting to fetch http://distcache.FreeBSD.org/ports-distfiles/arduino/arduino-Arduino-1.6.18-b439a77_GH0.tar.gz
arduino-Arduino-1.6.18-b439a77_GH0.tar.gz                0  B    0  Bps
=> arduino-ArduinoCore-sam-fc57f2e_GH0.tar.gz doesn't seem to exist in /portdistfiles/arduino.
=> Attempting to fetch https://codeload.github.com/arduino/ArduinoCore-sam/tar.gz/fc57f2e?dummy=/arduino-ArduinoCore-sam-fc57f2e_GH0.tar.gz
fetch: https://codeload.github.com/arduino/ArduinoCore-sam/tar.gz/fc57f2e?dummy=/arduino-ArduinoCore-sam-fc57f2e_GH0.tar.gz: size unknown
fetch: https://codeload.github.com/arduino/ArduinoCore-sam/tar.gz/fc57f2e?dummy=/arduino-ArduinoCore-sam-fc57f2e_GH0.tar.gz: size of remote file is not known
arduino-ArduinoCore-sam-fc57f2e_GH0.tar.gz               0  B    0  Bps
=> arduino-ArduinoCore-samd-1.6.12_GH0.tar.gz doesn't seem to exist in /portdistfiles/arduino.
=> Attempting to fetch https://codeload.github.com/arduino/ArduinoCore-samd/tar.gz/1.6.12?dummy=/arduino-ArduinoCore-samd-1.6.12_GH0.tar.gz
fetch: https://codeload.github.com/arduino/ArduinoCore-samd/tar.gz/1.6.12?dummy=/arduino-ArduinoCore-samd-1.6.12_GH0.tar.gz: size unknown
fetch: https://codeload.github.com/arduino/ArduinoCore-samd/tar.gz/1.6.12?dummy=/arduino-ArduinoCore-samd-1.6.12_GH0.tar.gz: size of remote file is not known
arduino-ArduinoCore-samd-1.6.12_GH0.tar.gz               0  B    0  Bps
=> Fetched file size mismatch (expected 3321356, actual 3323979)
=> Trying next site
=> Attempting to fetch http://distcache.FreeBSD.org/ports-distfiles/arduino/arduino-ArduinoCore-samd-1.6.12_GH0.tar.gz
fetch: http://distcache.FreeBSD.org/ports-distfiles/arduino/arduino-ArduinoCore-samd-1.6.12_GH0.tar.gz: Not Found
=> Couldn't fetch it - please try to retrieve this
=> port manually into /portdistfiles/arduino and try again.
*** Error code 1

Stop.
make: stopped in /usr/ports/devel/arduino-core
Comment 1 Nathan 2018-08-31 01:19:19 UTC
Just tested, and mine went fine; Took a bit for it to finally start downloading, but got no mis-match error
Comment 2 freebsd 2018-09-11 21:13:58 UTC
I just tried to download the arduino-Arduino-1.6.18-b439a77_GH0.tar.gz tarball from home, work and multiple co-lo facilities we have through work.  In each instance I was coming from a different subnet and different network space (different ARIN allocation).  The locations I was downloading form were Frankfurt Germany, London England, and four locations in the US.  Some requests came via proxies others were direct.

In all 5 cases the arduino-Arduino-1.6.18-b439a77_GH0.tar.gz file came back with a size of 48187647.  I've used both "fetch" and "wget" to get the file.  In all cases the MD5 checksum is 7de2b036c27cf31d97c3d01ef0a19a32 and the SHA256 checksum is 70191e9badd0986b697326d4640da1d24090f0c72a7e8c8f783bcb1ca63121a4

Below is the distinfo file for devel/arduino-core.  This is the one I have for my poudriere setup.  This is the same contents that I have from a fresh portsnap as well.
user@freebsd:/usr/local/poudriere/ports/local/devel/arduino-core
R cat distinfo 
TIMESTAMP = 1490895941
SHA256 (arduino/arduino-Arduino-1.6.18-b439a77_GH0.tar.gz) = ded6eebe1b58834df432e8d49f01a8c1aeb1328479f244075db7f76c12019f37
SIZE (arduino/arduino-Arduino-1.6.18-b439a77_GH0.tar.gz) = 48187661
SHA256 (arduino/arduino-ArduinoCore-sam-fc57f2e_GH0.tar.gz) = 3d374e7c7037d317f5da73c95361033b81e6cc12330a7623b20ad2f084bff247
SIZE (arduino/arduino-ArduinoCore-sam-fc57f2e_GH0.tar.gz) = 38839710
SHA256 (arduino/arduino-ArduinoCore-samd-1.6.12_GH0.tar.gz) = fefa0cb0871033451aa0ea08b6a9b4587acdd800b6064a9d47233be7b6368ce8
SIZE (arduino/arduino-ArduinoCore-samd-1.6.12_GH0.tar.gz) = 3321356

That timestamp is UTC: Thu Mar 30 17:45:41 2017


The DNS entries that all 5 locations report for codeload.github.com are:
$ host codeload.github.com
codeload.github.com has address 192.30.253.121
codeload.github.com has address 192.30.253.120

Each location has it's own DNS servers that are independent of each other.


I have downloaded files from both IPs just to validate that they are reporting the same (used wget enough times until both IPs were represented).

Being the maintainer is it possible that you have a locally modified ports tree that hasn't been synced up?
Comment 3 Kyle Evans freebsd_committer 2018-09-11 21:21:31 UTC
(In reply to freebsd from comment #2)

Hi,

Apologies for the silence; I'll take a look at this in a bit. Seems to be a case of github re-rolling the distfile, so I'll need to compare.

Thanks for the report!
Comment 4 Kyle Evans freebsd_committer 2018-09-12 02:40:05 UTC
Created attachment 197040 [details]
svn(1) diff against the ports tree

Hi,

I've verified with diffoscope that the change in the distfile is innocuous. Uploading a diff against the ports tree to re-roll distinfo, but I'm double-checking that devel/arduino18 isn't also affected before poking a ports committer to commit or approve me for commit.