[Spread-users] Reliable?

John Schultz jschultz at d-fusion.net
Fri Apr 11 14:08:07 EDT 2003


I re-read that post. It is mostly correct but there are two important 
things that he got wrong or over simplified.

The virtual synchrony property does not apply to non-Causal messages -- 
he incorrectly stated that if processes p and q move togther from view v 
to view w then they received the same set of messages in view v. This 
property is only true for CAUSAL and higher service lvl msgs.

For simplicity I think he skirted over how to compute if two processes 
moved together from v to w. The transitional membership set is what 
provides this information. Inspecting the membership sets of v and w and 
computing the intersection is NOT enough.

-- 
John Schultz
Co-Founder, Lead Engineer
D-Fusion, Inc. (http://www.d-fusion.net)
Phn: 443-838-2200 Fax: 707-885-1055

Ryan Caudy wrote:

>

> http://lists.spread.org/pipermail/spread-users/2003-April/001362.html 
> contains a pretty good discussion of this.
> 
> Cédric Coulon wrote:
> 
>> Hello,
>>
>> If a node falls during a reliable message is sent (FIFO_MESS for 
>> example),
>> are the other nodes received the message after all?
>>
>> Thanks.
>>  
>> ----
>> Cédric Coulon                                          Tel:   02 51 12 
>> 59 63
>> IRIN, 2, rue de la Houssiniere, BP 92208    Fax: 02 51 12 58 12
>> 44322 NANTES cedex 3
>> FRANCE                     mailto:Cedric.Coulon at irin.univ-nantes.fr
> 






More information about the Spread-users mailing list