[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: BUG Commit within samba-network dir

From: Stefan Küng <tortoisesvn_at_gmail.com>
Date: Wed, 27 Aug 2008 09:01:23 +0200

Ben Voigt wrote:
>> Have you considered that the change in TSVN may have exposed a problem
>> with Samba?
>>
>> If it works properly locally, and breaks on Samba, the problem must be
>> with Samba and/or the network. TSVN doesn't see the local and remote
>> paths differently.
>
> This ("the problem MUST be") is also false, for the same reasoning.
>
> "works locally" is not comprehensive enough testing to rule out bugs.
> For example, the difference in time taken to complete I/O operations
> could easily affect the behavior of a race condition. Overlapped I/O
> may complete in a different order over the network. Packet
> fragmentation can cause partial completion of operations. Virus scans
> could interact with network traffic. Lost packets could result in
> failed operations and trigger error handling and retry code that never
> gets exercised locally (this should not be dismissed as a "problem
> with the network") There are a LOT of other factors involved with
> accessing file shares beyond just Samba.

Not really.
Sure, the comparison between Samba and local harddrive access is not the
same and this could still be a bug in Subversion. But all your reasoning
goes away once you compare Samba shares and Windows shares: we never
heard of any problems with Windows network shares, only with Samba
shares. And those two have the same timing factors.

Stefan

-- 
       ___
  oo  // \\      "De Chelonian Mobile"
 (_,\/ \_/ \     TortoiseSVN
   \ \_/_\_/>    The coolest Interface to (Sub)Version Control
   /_/   \_\     http://tortoisesvn.net

Received on 2008-08-27 09:01:50 CEST

This is an archived mail posted to the TortoiseSVN Users mailing list.