FS#20022 - Download progress and completion issues during repo sync

Attached to Project: Pacman
Opened by trusktr (trusktr) - Wednesday, 30 June 2010, 06:28 GMT
Last edited by Allan McRae (Allan) - Sunday, 31 January 2016, 04:03 GMT
Task Type Bug Report
Category Backend/Core
Status Closed
Assigned To Dave Reisner (falconindy)
Architecture All
Severity Low
Priority Normal
Reported Version 3.4.0
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

Description:
Pacman shows package sync incorrectly. it appeared to sync all the repos, but then two of them disappeared so in the end it looked like this:

[root@rocketship ~]# pacman -Syu
:: Synchronizing package databases...
core is up to date
community 377.4K 1029.5K/s 00:00:00 [######################] 100%
archlinuxfr is up to date
arch-games is up to date
:: Starting full system upgrade...

^^^ notice core and extra are missing, although i saw progress bars for them at first.



Additional info:
* package version: pacman 3.4.0-2


Steps to reproduce:
Do pacman -Syu
This task depends upon

Closed by  Allan McRae (Allan)
Sunday, 31 January 2016, 04:03 GMT
Reason for closing:  Works for me
Additional comments about closing:  Appeared to be an issue with mirror round robin that has not been seen in a long time.
Comment by trusktr (trusktr) - Wednesday, 30 June 2010, 06:30 GMT
Formatted (monospaced) version of pacman output: http://pastebin.com/mEkM3KDK

This doesn't seem like a critical bug, but nonetheless, i noticed it and thought "hey, where's the other repos?"
Comment by trusktr (trusktr) - Wednesday, 30 June 2010, 06:32 GMT
Hey! Upon tryin pacman -Syu again, i got this:
http://pastebin.com/g1GGWMu1

So something happened (possibly interrupted pacman).
Comment by Allan McRae (Allan) - Wednesday, 30 June 2010, 09:14 GMT
  • Field changed: Attached to Project (Arch Linux → Pacman)
Please file pacman bugs in the pacman section
Comment by Allan McRae (Allan) - Wednesday, 30 June 2010, 09:16 GMT
@Dan: this seems very similar to what you observed prior to the 3.4 release.
Comment by Allan McRae (Allan) - Wednesday, 30 June 2010, 09:17 GMT
@trusktr: what mirror do you use?
Comment by Dan McGee (toofishes) - Wednesday, 30 June 2010, 12:03 GMT
Yes, this looks exactly like what I was seeing before, interesting. For reference:
http://mailman.archlinux.org/pipermail/pacman-dev/2010-June/011095.html
Comment by trusktr (trusktr) - Wednesday, 30 June 2010, 22:33 GMT
Hey guys, this is the ranked server list i'm using:
Server = http://mirrors.kernel.org/archlinux/$repo/os/x86_64
Server = ftp://mirrors.kernel.org/archlinux/$repo/os/x86_64
Server = http://mirror.leaseweb.com/archlinux/$repo/os/x86_64
Server = http://ftp.yz.yamagata-u.ac.jp/pub/linux/archlinux/$repo/os/x86_64
Server = http://archlinux.puzzle.ch/$repo/os/x86_64
Server = http://mirror.aarnet.edu.au/pub/archlinux/$repo/os/x86_64
Server = http://pet.inf.ufsc.br/mirrors/archlinux/$repo/os/x86_64
Server = http://cesium.di.uminho.pt/pub/archlinux/$repo/os/x86_64
Server = ftp://ftp.free.fr/mirrors/ftp.archlinux.org/$repo/os/x86_64
Server = ftp://cesium.di.uminho.pt/pub/archlinux/$repo/os/x86_64
Comment by Allan McRae (Allan) - Thursday, 01 July 2010, 00:46 GMT
Hmmm... same top mirror entry... May be a coincidence or perhaps something with kernel.org's geolocationing?
Comment by trusktr (trusktr) - Thursday, 01 July 2010, 01:29 GMT
So what do you think is possible regarding geolocationing? That possibly a byte or two got corrupted during download?
Comment by Dan McGee (toofishes) - Friday, 10 September 2010, 22:03 GMT
Reproduced again today, got some debug logging which is attached.
   weird-download-extra.txt (5.8 KiB)
Comment by Allan McRae (Allan) - Sunday, 12 September 2010, 10:08 GMT
Similar: https://bbs.archlinux.org/viewtopic.php?id=104766 Also kernel.org...
Comment by Allan McRae (Allan) - Thursday, 24 March 2011, 03:00 GMT
debug: ust.mtime: 1284151607 local_time: 1284143674 compare: -7933
debug: ust.size: 474076 local_size: 474076 compare: 0
...
error: extra.db.tar.gz appears to be truncated: 474047/474076 bytes

So... times are different but the sizes are the same - seems unlikely. And then the size is wrong. Is that information obtained in a single call? Or could it be possible that two different "kernel.org" mirrors that are out of sync are being queried?
Comment by Allan McRae (Allan) - Saturday, 04 August 2012, 08:27 GMT
@Dave/Dan: can we close this? I have not seen this issue in a long time... It was almost certainly due to getting different servers from a round robin system, which I am not sure we can do anything about.

Loading...