Nope. Our default window size is about 8K. The socket is probably about the same. If an error closes the socket before you've finished writing, you're likely to see a write error rather than a polite error message.
Static gifs and docs. Perhaps it would have been better to protect them against unauthorized access, but since anyone who wants them can get them by purchasing a cache and we don't view the docs as a profit center, it didn't strike us as a high priority.
If we had any secret scripts accessible as disk_objects, this might be a problem. Since we don't, this is merely a clever way to confuse your browser.
Our code is not perfect, but the absolute opinions expressed are absolutely wrong.
--bob-- Cache breaks, compiling Ash leaves shimmer beneath clouds, Nerf ball strikes face.
I think any security issue is high priority to fix especially if issue has been posted to bugtrack. It hardly inspires confidence that serious security issues will be dealth with in the correct and proper manner.
This is joke especially since a simple strings of the code base shows there are undocumented code features.
Colin Johnston SA PSINET UK
In message Pine.GSO.4.05.10003311053290.12018-100000@staff.uk.psi.com, Colin Johnston writes:
Let me put it another way. The only things in the disk object directory are static gifs and administration documentation. The only security risk they present is that someone who hasn't purchased a cache can read them. Since the same documents are available on any netcache, anywhere, denying access to them will not improve system security.
There are undocumented features, but there are no scripts in the disk_object directory. I don't know whether there are undocumented documents.
--bob-- Cache breaks, compiling Ash leaves shimmer beneath clouds, Nerf ball strikes face.
Hi, I seem to have found a log rotation bug in NetCache Appliance 4.0 R4 code release. It seems the messages file is not rotating anymore and this is creating huge 2mb+ messages files so when weekly autosupport emails are sent large emails are being produced.
Does anyone know if this is fixed in 4.1R1 code release and if not can it be fixed for the next early code release ??
Thanks
Colin Johnston SA PSINET UK
* Colin Johnston (colinj@uk.psi.com) done spit this rhetoric:
Howdy Colin!
Are there any messages regarding log rotation being produced when the log should be rotated?
-
Colin Johnston -
Jamie Lawrence -
Jerry Talkington -
Robert English