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

RE: use of plaintext passwords in svnserve.conf and client-side caches

From: Méresse Christophe <christophe.meresse_at_nagra.com>
Date: 2006-10-18 14:49:55 CEST

> -----Original Message-----
> From: news [mailto:news@sea.gmane.org]On Behalf Of Alexis Huxley
> Sent: mercredi, 18. octobre 2006 13:08
> To: users@subversion.tigris.org
> Subject: use of plaintext passwords in svnserve.conf and client-side
> caches
[...]
> (I am aware of svn+ssh://, but am currently at the stage of a
> migration
> project where I want to try as many access methods as possible in
> order to collect some performance data before deciding which protocols
> will be made available to users.)

Hello,

This does not give an answer to your question but here is
a bit of my experience.

svn:

----
pros - quick
cons - problem of the plaintext passwords (No way to manage 
       these files from the NIS or another authentication tool (?))
svn+ssh:
--------
pros - uses the NIS authentication
cons - the ssh does not modify the performance on one command 
       but is extremely annoying for the graphical tools or scripts 
       that need to do multiple successive commands (the ssh tunnel 
       creation time takes some tenths of a seconde each time that 
       can lead to many seconds or even minutes (that's the case 
       for tkCVS/tkSVN or tortoise))
     - We did not found how to avoid to give the read-write access 
       directly on the repository database as we have to allow the 
       ssh connection to the server (I would be very interested to know...) 
http:
-----
pros - quick
     - uses the apache authentication (NIS, PAM...)
     - opens interesting http usages !
cons - ? (relying on more layers, not encrypted)
https:
------
pros - quick (the ssh performance problem does not seem to appear with ssl)
     - uses the apache authentication (NIS, PAM...)
     - secure
     - opens interesting http usages !
cons - ? (relying on more layers)
A checkout is anyway about 4 times slower with Subversion than with CVS 
(Probably due to the creation of the subversion workspaces that contain 
many more files than CVS's ones)
I would be very interested to get the different feedbacks, experiences 
or advices about all these performance issues.
Regards,
Christophe Méresse
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Wed Oct 18 14:50:51 2006

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

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.