[cvsnt] NAV file system realtime protection, repository corruption, CVS version????

[Tony Hoyle]

On Wed, 15 Jan 2003 09:19:39 -0700, "Darrell Werries"
<dwerries at coretronics.com> wrote:

>
>1. Is this corruption isolated to the new version?  We have been running
>build 57d for about 6 months with this NAV feature turned on, and we have
>not seen any corruption in our repository...at least, as far as I know.
>Maybe we have been lucky?  Maybe it has occurred, but I have not come across
>it, yet?  How can I check?

No, it's random.  There have been reports going back 18 months that I
can remember.

>2. Are there some specifics as to what does occur and exactly what causes it
>to occur?  Can/should I monitor the cause within NAV or the effect within
>our repository?  Symptoms?  Automated monitoring?  Automated cleanup?  Etc.?

No idea...  There's be no feedback from Norton on this and those that
have tried to replicate it have had limited success.

>3. I could turn this feature off in NAV, but it sure does seem like a rather
>important thing to be checking for, particularly on a server (network,
>email, internet, web, etc.).

Not really...  If you guard the entrances and exits to your network -
virus check the mail server, ftp, etc.  and have policies that forbid
downloading of kazaa etc. you'll be OK (for those that insist, a
desktop AV package is more than adequate to scan downloaded files).
The file protection feature doesn't seem to do anything particularly
useful - especially on a cvs server that doesn't actually do anything
except store source files etc.  There's simply no way a virus could
get on there (and no way an AV package could see it anyway as
executables like everything else are stored in RCS files).

>4. Reasons to upgrade or not to upgrade from 57d -> 63?

Bug fixes, mostly.  If you haven't had any problems & don't need the
new features then there's little advantage.

Tony