About this list Date view Thread view Subject view Author view Attachment view

From: Herbert Poetzl (herbert_at_13thfloor.at)
Date: Tue 19 Jul 2005 - 23:16:20 BST


On Tue, Jul 19, 2005 at 07:32:17PM +0200, Nicolas Costes wrote:
> Le mardi 19 Juillet 2005 07:33, Herbert Poetzl a écrit :
> > it will take a few more days (maybe a fortnight) until
> > we are convinced that it is _final_ and 2.0 will then
> > be release (of course, shortly after that release, a lot
> > of bugs/issues will be discovered, when the folks waiting
> > for a stable release start using it, instead of testing
> > the release candidates :)
>
> (Lol.)
>
> I'd like to help by testing the last rc (At home, it's holiday time
> here). But the problem is that apart from creating a vserver, starting
> it and applause, I don't really know what can be done to _really_ test
> it...

basically everybody has a slightly different setup and
different usage for linux-vserver so 'just' testing it
already provides valuable information, especially if
something doesn't work as expected ... (although we
always prefer reports stating "It just works great!")

> Is there a check-list somewhere ?

basically the 'feature' list on the Release FAQ page
(http://linux-vserver.org/Release+FAQ) gives a good
idea what could/should be tested ...
of course, previous issues/incidents or security in
general might be a good idea to test too (you can't
really overdo it here :)

> Some specific torture tests you'd like to be run ?

benchmarks, stress tests (inside and outside a guest)
special conformance tests (like for example the LTP)
but the important part is to classify the results and
feed back the 'distilled' information of interest to
the developers (or make some fancy marketing paper
like the one showing that Xen really rocks :)

> Would it be interresting to create a "torture"
> documentation on the wiki ?

of course, and even better would be to create some
kind of linux-vserver test project (for performance
and regression testing), that would definitely improve
linux-vserver in the long run ...

btw, I already started some regression testsing with
the testme.sh and testfs.sh scripts (which focus on
general functionality and xid/xattr tests)

very important (but until now not really covered) areas
would IMHO be:

 - limits and accounting
 - process and resource isolation
 - virtualization
 - scheduler testing

best,
Herbert

> --
> Brigandage et truandage politique sont
> les deux gamelles de la France. SIM

> _______________________________________________
> Vserver mailing list
> Vserver_at_list.linux-vserver.org
> http://list.linux-vserver.org/mailman/listinfo/vserver

_______________________________________________
Vserver mailing list
Vserver_at_list.linux-vserver.org
http://list.linux-vserver.org/mailman/listinfo/vserver


About this list Date view Thread view Subject view Author view Attachment view
[Next/Previous Months] [Main vserver Project Homepage] [Howto Subscribe/Unsubscribe] [Paul Sladen's vserver stuff]
Generated on Tue 19 Jul 2005 - 23:16:44 BST by hypermail 2.1.3