IHAC that needs proof that traffic between local zones on the same system
will not in fact hit the NIC card. There's two scenarios.
First, if one does FTP or scp to transfer files between the zones,
how can I prove it doesn't hit the NIC or wire? What function in
the IP driver would I look for
Brian Kolaci wrote:
IHAC that needs proof that traffic between local zones on the same system
will not in fact hit the NIC card. There's two scenarios.
An interesting demand. I wonder if this person has ever asked for proof of other
proper computer operation. Do they demand that (pick your
Jeff Victor wrote:
Brian Kolaci wrote:
IHAC that needs proof that traffic between local zones on the same
system
will not in fact hit the NIC card. There's two scenarios.
An interesting demand. I wonder if this person has ever asked for
proof of other proper computer operation. Do th
Brian Kolaci writes:
> IHAC that needs proof that traffic between local zones on the same system
> will not in fact hit the NIC card. There's two scenarios.
I guess the first question is what sort of proof would be
satisfactory.
If we can cite parts of the code, would that help? Or does this
pe
Brian Kolaci writes:
> > How is 'snoop' insufficient?
> >
> The question came up as to whether snoop sees all NIC traffic or just
> wire traffic.
> He noted that some NIC's still process the packet but just pass it back
> to the
> driver rather than the wire... He wants to prove that it doesn't
Hi:
DTrace is the tool instrumenting the request.
Thanks
Lei
- Original Message -
From: James Carlson <[EMAIL PROTECTED]>
Date: Wednesday, July 26, 2006 9:36 am
Subject: Re: [zones-discuss] traffic/data between zones
To: Brian Kolaci <[EMAIL PROTECTED]>
Cc: zones-discuss@opensolaris.org
Lei Liu writes:
> DTrace is the tool instrumenting the request.
[...]
> > I would check for calls to put(9F) and putnext(9F) from within the IP
> > module (which includes TCP).
... then those are the calls you want.
--
James Carlson, KISS Network<[EMAIL PROTECTED]>
Sun Micros
James Carlson wrote:
Brian Kolaci writes:
How is 'snoop' insufficient?
The question came up as to whether snoop sees all NIC traffic or just
wire traffic.
He noted that some NIC's still process the packet but just pass it back
to the
driver rather than the
Brian Kolaci wrote:
Essentially he wants to be able to assure his users that the transfers
will be fast and not affect the network users. If they complain that
during these operations are slow, then he wanted to recommend sharing
a lofs mount and share the data that way (and also be assured tha
With the performance boosts included in recent solaris versions I'm
told that there's not much of a difference between handing the database
raw devices vs. using a filesystem anymore.
To test this out, my customer would like to try both ufs and vxfs
filesystems in the global zone and lofs mount
On Wed, 26 Jul 2006, Brian Kolaci wrote:
> To test this out, my customer would like to try both ufs and vxfs
> filesystems in the global zone and lofs mount them to a local zone
> and test the database on that lofs mount.
I'd also recommend that your customer tries the tests on ZFS too.
--
Rich
>The question came up as to whether snoop sees all NIC traffic or just
>wire traffic.
>He noted that some NIC's still process the packet but just pass it back
>to the
>driver rather than the wire... He wants to prove that it doesn't even
>hit the NIC.
>Will snoop detect all traffic to/from the
Is it possible to add an inherit-pkg-dir to a zone's config after "zoneadm ...
install"? I'm confident the answer is 'no' but just had to ask.
The man page and docs.sun.com are a little vague on this. They both say "Such
resources [those pkgs] are not modifiable or removable once a zone has be
Jeff Victor wrote:
Is it possible to add an inherit-pkg-dir to a zone's config after
"zoneadm ... install"? I'm confident the answer is 'no' but just had to
ask.
The man page and docs.sun.com are a little vague on this. They both say
"Such resources [those pkgs] are not modifiable or removab
14 matches
Mail list logo
