Comments on: RHEL 5.3 is out

By: ngm

ngm — Wed, 18 Feb 2009 09:42:32 +0000

If you really need the latest php/apache version, you can download and rebuild the srpm from fedora.

By: Luigi Rosa

Luigi Rosa — Wed, 11 Feb 2009 10:30:55 +0000

PHP 5.2
I really miss it.

By: Ed

Ed — Wed, 04 Feb 2009 17:40:38 +0000

No updates to packages that SOOO badly need it (php and apache come to mind, amongst piles of others)

The fun of adding ext4 (sorry ext4dev!) but still no xfs?

upgrading our rhel 5.2->5.3 left it not booting. grub didn’t get the new kernel name added.
a number of small depsolves every day since.

On the plus side Oracle runs even better now than before.

By: Vygantas

Vygantas — Mon, 02 Feb 2009 20:32:34 +0000

ext4 for sure

By: Jim Perrin

Jim Perrin — Mon, 02 Feb 2009 16:29:11 +0000

It’s usually between 2 weeks and a month. The rebuild process takes a while, then QA gets it and works on it for a while. Then it gets seeded to the primary mirrors. Once enough mirrors have the release it gets announced.

By: Justin

Justin — Sat, 31 Jan 2009 17:52:38 +0000

I understand that there’s no hard and fast date on which CentOS follows a RH release, but what’s the typical timeframe? Weeks? Months? I need to rebuild a box with 5.2 on it, and I’m wondering whether I should wait for 5.3 to drop.

By: Jim Perrin

Jim Perrin — Fri, 30 Jan 2009 23:16:29 +0000

Interesting. I’m seeing a different problem from our DS4700. while I’m not installing directly to the san, configuring the multipath bits after install has led to some interesting issues. Using the multipath defaults config for the 1815, modified for the 1814 ds4700 value, Everything appears normal unless disk i/o is called for. For whatever reason, the i/o takes place on the wrong path every time. As soon as the i/o operation is finished, the storage manager java client shows the ds4700 as healthy again. Every IO operation causes it to generate an error state and show that the drive is no longer on the preferred path. Our upstream ticket with redhat has so far yielded nothing beyond a ‘not all ds4xxx chassis are supported’. The ticket is still open, and I plan on escalating it as far as I can.

By: Jacques

Jacques — Thu, 29 Jan 2009 09:24:14 +0000

Well, unfortunately the IBM DS4800 support seems to be broken…. We have upstream vendor support, and I’ve tried doing a “linux mpath” install using a blade connected to a DS4800, and the installer breaks before even getting to the disk partitioning, sometimes it gets there, sometimes it ever tries to install some packages, but it always breaks, saying the disk is read-only.

I’ve logged a call with the upstream vendor, waiting for feedback still.

By: Shankar

Shankar — Mon, 26 Jan 2009 23:56:09 +0000

Unfortunately, they didn’t refresh unixODBC in this release. The current one (2.2.11) doesn’t allow Oracle’s 10.2.0.4 and 11.0 ODBC drivers to be loaded, because they depend on a new symbol that’s only in 2.2.12.

Sigh!

By: cheguaka

cheguaka — Mon, 26 Jan 2009 12:16:59 +0000

iwl3945

Yes. I use CentOS at my laptop.