Use rpm2tgz to unpack the sqlite RPM on another system, then unpack the
resulting tarball on your affected system. This will drop the sqlite
binaries into place, should at least get RPM working again.
Paul Krizak 7171 Southwest Pkwy MS B200.3A
MTS Systems Engineer
Sorry that's rpm2cpio...
Paul Krizak 7171 Southwest Pkwy MS B200.3A
MTS Systems EngineerAustin, TX 78735
Advanced Micro Devices Desk: (512) 602-8775
Linux/Unix Systems Engineering Cell: (512) 791-0686
Global IT Infrastructure
I don't know the ETA for the fix to show up everywhere, but I do know
there is a fix (bz 725536), because RH Support was able to get us a test
package that fixes it.
-Paul
On 08/24/11 05:36, Horst Severini wrote:
Speaking of known bugs: is there an ETA for the fix to the autofs bug
which
Sounds like you need a block-level solution like NetApp's SnapMirror. I
don't think trying to synchronize this data from the OS layer is going
to be a very elegant or efficient solution.
Paul Krizak 7171 Southwest Pkwy MS B200.3A
MTS Systems Engineer
stepping: 3
cpu MHz : 1000.000
The Opteron 1222 is a 3.0Ghz processor -- it runs at 1.0Ghz when idle.
Paul Krizak 7171 Southwest Pkwy MS B200.3A
MTS Systems EngineerAustin, TX 78735
Advanced Micro Devices Desk: (512) 602-8775
, it may disable 4 of the
cylinders, so your cpu MHz in this example would be V4 3.2L even
though the model shows up as a V8 6.4L.
Paul Krizak 7171 Southwest Pkwy MS B200.3A
MTS Systems EngineerAustin, TX 78735
Advanced Micro Devices Desk: (512
Paul Krizak 7171 Southwest Pkwy MS B200.3A
MTS Systems EngineerAustin, TX 78735
Advanced Micro Devices Desk: (512) 602-8775
Linux/Unix Systems Engineering Cell: (512) 791-0686
Global IT InfrastructureFax: (512) 602-0468
performance out of
G7's.
Paul Krizak 7171 Southwest Pkwy MS B200.3A
MTS Systems EngineerAustin, TX 78735
Advanced Micro Devices Desk: (512) 602-8775
Linux/Unix Systems Engineering Cell: (512) 791-0686
Global IT Infrastructure
are detected (and not just the first 4).
HP is pretty good about updating the PSP when new hardware comes out, so
I imagine that v8.6 probably supports the 585 G7 just fine.
Paul Krizak 7171 Southwest Pkwy MS B200.3A
MTS Systems EngineerAustin, TX 78735
We're looking at it in our environment too. Very interested in any
experience good/bad.
Paul Krizak 7171 Southwest Pkwy MS B200.3A
MTS Systems EngineerAustin, TX 78735
Advanced Micro Devices Desk: (512) 602-8775
Linux/Unix Systems
, but in our case it
would satisfy our auditing tool, which runs in userland as a script (and
thus doesn't make uname() system calls directly).
Have you had any experience with the kSplice API?
Paul Krizak 7171 Southwest Pkwy MS B200.3A
MTS Systems Engineer
Fantastic feedback. Thanks!
Paul Krizak 7171 Southwest Pkwy MS B200.3A
MTS Systems EngineerAustin, TX 78735
Advanced Micro Devices Desk: (512) 602-8775
Linux/Unix Systems Engineering Cell: (512) 791-0686
Global IT Infrastructure
I'm also curious about the PSP since we use it heavily in our
infrastructure.
Paul Krizak 7171 Southwest Pkwy MS B200.3A
MTS Systems EngineerAustin, TX 78735
Advanced Micro Devices Desk: (512) 602-8775
Linux/Unix Systems Engineering
Yes, you can run them out of the box with RHEL5.5.
Paul Krizak 7171 Southwest Pkwy MS B200.3A
MTS Systems EngineerAustin, TX 78735
Advanced Micro Devices Desk: (512) 602-8775
Linux/Unix Systems Engineering Cell: (512) 791-0686
Global
Seconded. I'm only getting 4.5KiB/sec...
Paul Krizak 7171 Southwest Pkwy MS B200.3A
Senior Systems Engineer Austin, TX 78735
Advanced Micro Devices Desk: (512) 602-8775
Linux/Unix Systems Engineering Cell: (512) 791-0686
Global
The new 8/12-core Magny-Cours stuff (G34 socket) is going to be under
the 6000-series, with the 4/6-core Lisbon stuff (C32 socket) under the
4000-series.
Paul Krizak 7171 Southwest Pkwy MS B200.3A
Senior Systems Engineer Austin, TX 78735
Advanced Micro
Yes, it's the Mentor Calibre, and it's multi-threaded, so it's able to
take advantage of lots of cores. I was running a Calibre-based
benchmark when I took this screenshot.
Paul Krizak 7171 Southwest Pkwy MS B200.3A
Senior Systems Engineer Austin, TX 78735
-height/2-socket blade, or 48 cores in a full-height/4-socket blade.
Paul Krizak 7171 Southwest Pkwy MS B200.3A
Senior Systems Engineer Austin, TX 78735
Advanced Micro Devices Desk: (512) 602-8775
Linux/Unix Systems Engineering Cell: (512) 791
We're still using RHEL4...so RHEL5 is still new for us :-)
Paul Krizak 7171 Southwest Pkwy MS B200.3A
Senior Systems Engineer Austin, TX 78735
Advanced Micro Devices Desk: (512) 602-8775
Linux/Unix Systems Engineering Cell: (512) 791-0686
initrd.img, in pxelinux.cfg, and make sure that the initrd.img is the
one from the original ISOs under images/pxeboot.
Paul Krizak 7171 Southwest Pkwy MS B200.3A
Senior Systems Engineer Austin, TX 78735
Advanced Micro Devices Desk: (512) 602-8775
At one time, we used to have to put apm=power-off on the kernel
command line to get systems to reboot. Not sure if that's still
something. You could try it though...
Paul Krizak 7171 Southwest Pkwy MS B200.3A
Senior Systems Engineer Austin, TX 78735
had to be able to bring up
eth1 on a private network to connect to the iSCSI shelf via a different
switch. We never tried using the primary interface to do this, since in
our environment, SAN traffic is not allowed to share the network with
the other traffic.
Paul Krizak
a locally-attached disk or a FC SAN or something.
Paul Krizak 7171 Southwest Pkwy MS B400.2B
Senior Systems Engineer Austin, TX 78735
Advanced Micro Devices Desk: (512) 602-8775
Linux/Unix Systems Engineering Cell: (512) 791-0686
Silicon
is a pretty easy win.
Paul Krizak 7171 Southwest Pkwy MS B400.2B
Senior Systems Engineer Austin, TX 78735
Advanced Micro Devices Desk: (512) 602-8775
Linux/Unix Systems Engineering Cell: (512) 791-0686
Silicon Design Division Fax
Our testing shows that RHEL5's kernel finally handles NUMA properly.
Leave the ACPI SRAT table enabled, and disable node interleaving for
best performance.
RHEL3 and RHEL4 (up to U5) still need node interleaving ON for best
performance. RHEL4 U6+ seems to be better with NUMA.
Paul Krizak
, then kickstart can fetch that and use it.
The benefit is that you can use any NIC in the system that supports PXE.
You configure the BIOS to boot from the NIC you want, and pxelinux and
kickstart will take care of the rest.
Paul Krizak 7171 Southwest Pkwy MS B400.2A
Advanced Micro
provides should be served
by domU's, not Dom0. This is because anybody that can gain access to
your Dom0 can manipulate your DomU's, which is a huge security problem.
Paul Krizak 7171 Southwest Pkwy MS B400.2A
Advanced Micro Devices Austin, TX 78735
Linux/Unix
You can specify network info on the kernel boot line, i.e.:
linux ks=http://blah/ks.cfg ip=$ip netmask=$netmask gateway=$gateway
dns=$dns_server
Paul Krizak 7171 Southwest Pkwy MS B400.2A
Advanced Micro Devices Austin, TX 78735
Linux/Unix Systems
-to-maintain hands-free installing.
If you don't have a DHCP server that you can configure for PXE and TFTP,
you may not be able to do this, though. You're already heading in the
right direction using a network-accessible kickstart file...
Paul Krizak 7171 Southwest Pkwy MS
In 5.2, it does not. This works instead:
[EMAIL PROTECTED] ~]$ grep Rgb /etc/X11/xorg.conf
RgbPath/usr/share/X11/rgb
Paul Krizak 7171 Southwest Pkwy MS B400.2A
Advanced Micro Devices Austin, TX 78735
Linux/Unix Systems Engineering Desk: (512) 602
suggest
that you play with the clients a bit -- it might not be a problem with
your NFS server.
Paul Krizak 7171 Southwest Pkwy MS B400.2A
Advanced Micro Devices Austin, TX 78735
Linux/Unix Systems Engineering Desk: (512) 602-8775
Silicon Design
Wow! I'm surprised that our TAM did not dig up these bugzilla entries.
Both of these definitely look like they could be causing (or at least
contributing to) our RHEL5 NFS performance woes.
Thanks for the heads-up!
Paul Krizak 7171 Southwest Pkwy MS B400.2A
Advanced
code and data is via the web,
which is how you're intending the data to be manipulated anyway.
Paul Krizak 7171 Southwest Pkwy MS B400.2A
Advanced Micro Devices Austin, TX 78735
Linux/Unix Systems Engineering Desk: (512) 602-8775
Silicon Design Division
. But 5.0 was pretty sketchy, IMHO. NFS/autofs
problems, Xen problems, etc. I'm avoiding 5.0 and starting out with 5.1
directly (with the updated -53.1.4 kernel).
Paul Krizak 5900 E. Ben White Blvd. MS 625
Advanced Micro Devices Austin, TX 78741
Linux
mpd.mhdc.trusted_hosts mpd.svdc.trusted_hosts
And so on and so on.
You sure the problem is with hosts.equiv/.rhosts? You might be facing a
PAM issue...
Paul Krizak 5900 E. Ben White Blvd. MS 625
Advanced Micro Devices Austin, TX 78741
Linux/Unix Systems Engineering
Have you checked /etc/security/access.conf? That's the one that always
screws us up...
Paul Krizak 5900 E. Ben White Blvd. MS 625
Advanced Micro Devices Austin, TX 78741
Linux/Unix Systems Engineering Phone: (512) 602-8775
Silicon Design Division
(obviously avoid things like /dev, /proc and /boot). What you
really want is a new system that is functionally identical to the old
one. No magic migration has to be done -- just make sure that all the
apps and config files and startup scripts are copied over to the new system.
Paul Krizak
at it!).
I've also noticed that the package *installation* takes much longer than
before. In RHEL3, it takes about 10 minutes to install 4GB of packages.
In RHEL5, it takes about an hour to install 7GB of packages. That
just ain't right.
Paul Krizak 5900 E. Ben White
You are correct -- I misspoke about this issue -- the permissions as set
on the server are unrelated to this issue (i.e. permissions set on the
client).
Paul Krizak 5900 E. Ben White Blvd. MS 625
Advanced Micro Devices Austin, TX 78741
Linux/Unix Systems
Yeah these are 8-64GB machines.
Paul Krizak 5900 E. Ben White Blvd. MS 625
Advanced Micro Devices Austin, TX 78741
Linux/Unix Systems Engineering Phone: (512) 602-8775
Silicon Design Division Cell: (512) 791-0686
Jay Turner wrote:
On Thu
Wow...running with an HTTP-based install cut the time down to 30 minutes
(the expected time for 10G of packages)
Looks like the NFS code in the PXE kernel is quite b0rked.
Paul Krizak 5900 E. Ben White Blvd. MS 625
Advanced Micro Devices Austin, TX 78741
between 20 and 70% CPU doing *something*, but no
installation is actually occurring.
Paul Krizak 5900 E. Ben White Blvd. MS 625
Advanced Micro Devices Austin, TX 78741
Linux/Unix Systems Engineering Phone: (512) 602-8775
Silicon Design Division
with the HTTP method and it's hung on this same package
each time.
How do I forcibly exclude an individual package without messing up the
dependency resolution? If it's a bum package, I should submit a bug to
redhat.
Paul Krizak 5900 E. Ben White Blvd. MS 625
Advanced
/gdm/defaults.conf
as a valid option...
--
Paul Krizak 5900 E. Ben White Blvd. MS 625
Advanced Micro Devices Austin, TX 78741
Linux/Unix Systems Engineering Phone: (512) 602-8775
Silicon Design Division Cell: (512) 791-0686
I've submitted bug #218133 about this -- looks like the RHEL5 theme is
what's broken, not gdm.
Paul Krizak 5900 E. Ben White Blvd. MS 625
Advanced Micro Devices Austin, TX 78741
Linux/Unix Systems Engineering Phone: (512) 602-8775
Silicon Design
::sigh:: naturally, I succeeded in answering my own question just
minutes after getting frustrated enough to send this e-mail.
Removing the console=tty0,9600 from the kernel boot line fixed it
right up.
Paul Krizak 5900 E. Ben White Blvd. MS 625
Advanced Micro Devices
46 matches
Mail list logo