[EMAIL PROTECTED] wrote:
first thing to say is that I like your new design ! I think it is good that:
So looks like the time was a good investment ;)
A couple of questions... 1. In the "Disadvantages" section you say that there will be larger exports. Is this really the case ? I would have thought that if you keep your CA and RA nodes in step (i.e. keep alternating import and exports) then the volume of the data exchange fil will be quite small and only contain items that have changed. Am I missing something ?
There is no step actually which checks that the objecttype and status must be exported to the other server. I simply export all changed stuff. If we add a simple check that checks the type and state then the exports are not larger if the servers are in step. The check is really simple (only one hash access) but I don't like to optimize before the basic design is commited.
2. Are you thinking of a seperate data exchange file per node ? Or a single data exchange file specific to each node ? If it is the former, then are you also thinking of a "select the node you wish to export to" type question during export ?
Hmm, this is very close to my English understanding limits :) Let's try it. First query yes. Second query no. Third query yes. The most installations only have two servers - online and offline. Therefore the most people have never to answer the question. If we have more servers then the people must know what they are doing.
If we would have only one single file then the size of the export file would be potentially much larger because we have to export every change. The select query would be in this case really simple: "select * from DX" and the update query would be reduced too - "update DX set NUMBER=EX where NUMBER=0". Otherwise is a such simple mechanism really alluring (I hope it is the right word).
3. I think you have an error in "3. RA", it says "- DELETE FROM DX WHERE SERVER_ID=0 and NUMBER>0 and NUMBER>=IX" where I think it should say "- DELETE FROM DX WHERE SERVER_ID=0 and NUMBER>0 and NUMBER<=IX". Are you testing us ? ;-)
No, I was only to fast because OpenSSL 0.9.8 does not compile cleanly. "./config shared" works perfectyl with our CVS HEAD but "./config shared no-deprecated" does not compile. I don't know what's the problem - OpenSSL or my options.
Michael -- _______________________________________________________________ Michael Bell Humboldt-Universitaet zu Berlin Tel.: +49 (0)30-2093 2482 ZE Computer- und Medienservice Fax: +49 (0)30-2093 2704 Unter den Linden 6 [EMAIL PROTECTED] D-10099 Berlin _______________________________________________________________
smime.p7s
Description: S/MIME Cryptographic Signature
