Re: Encryption protection

2001-03-02 Thread Kevin M. Myer

On Wed, 28 Feb 2001, Todd Reed wrote:

 I understand from the replies here that SimpleCrypt isn't secure, at
 least in the sense that with enough time the encryption scheme can be
 defeated. That's true for any scheme if you have infinite amounts of
 time and computers.

 What I'd like to know is a realistic assessment of its insecurity.
 Dantz is saying it is secure enough for the majority of commercial
 uses. Is the average script kiddie going to find SimpleCrypt easy t

I wouldn't believe that claim unless I saw the source for their algorithm
(not that I could review it but a thorough peer review would be necessary
to say the least to backup the claim - no pun intended).  Since 56-bit DES
is pretty easy to crack (it took all of three days on distributed.net a
few years ago and dedicated hardware is now available for that job), and
SimpleCrypt is an unpublished proprietary algorithm, it logically follows
that SimpleCrypt is also simpler to crack.

I'm curious about the DES encryption actually - it doesn't seem to matter
if I have encryption turned on or off - backups seem to take approximately
the same amount of time.  What exactly is being backed up:  the data
stream between client and server, the contents on the tape, the catalog
(on disk) or the header on the tape (or any combination thereof)?

   Basically, Retrospect's SimpleCrypt encryption method is faster than DES,
   but the tradeoff for speed yields a less robust encryption scheme.

Which basically says, to me, that its not very strong at all :)

Kevin
-- 
Kevin M. Myer
Systems Administrator
Lancaster-Lebanon Intermediate Unit 13
(717)-560-6140



--
--
To subscribe:[EMAIL PROTECTED]
To unsubscribe:  [EMAIL PROTECTED]
Archives:http://list.working-dogs.com/lists/retro-talk/
Search:  http://www.mail-archive.com/retro-talk%40latchkey.com/

For urgent issues, please contact Dantz technical support directly at
[EMAIL PROTECTED] or 925.253.3050.



Anyone with working multicast/broadcast setups?

2001-01-15 Thread Kevin M. Myer

Hi,

We have been moving some servers around here and one of the services that
got moved to a new machine (and a new subnet) is our Retrospect backup
software.  One of the immediate problems I found was that I could no
longer see the majority of my clients from the server.  [Actually this
problem existed before but since the server was originally on the subnet
that most of the clients were, it wasn't hard to enter the remaining
clients by address - however with DHCP and the addition of remote sites,
this becomes unmanageable].

I have 7 class C subnets that I wish to search for clients in.  The LAN is
a switched network with mostly Cisco Catalyst 2924s.  We have a Cisco 8540
layer 3 switch in place as well.  The problem is that, for good reason, we
have the layer 3 switch to not forward directed broadcasts to prevent
smurf attacks - and thats exactly one of the mechanisms Retrospect uses to
discover clients - a directed broadcast to the subnet.  We tried to get
the multicast stuff working via some configuration changes on the layer 3
switch but thats going to take some more research.

So I'm wondering if anyone has Retrospect working in a multiple subnet
environment with Cisco equipment AND with some reasonable security
precautions in place (like "no ip directed-broadcast").  Better yet,
anyone have it working with multicast, as this would be a preferred way to
do this I think.

For reference, clients are mostly Macs running 4.3 over TCP/IP, the server
is also a Mac running 4.3.

Thanks,

Kevin

-- 
Kevin M. Myer
Systems Administrator
Lancaster-Lebanon Intermediate Unit 13
(717)-560-6140



--
--
To subscribe:[EMAIL PROTECTED]
To unsubscribe:  [EMAIL PROTECTED]
Archives:http://list.working-dogs.com/lists/retro-talk/
Search:  http://www.mail-archive.com/retro-talk%40latchkey.com/

For urgent issues, please contact Dantz technical support directly at
[EMAIL PROTECTED] or 925.253.3050.