I hadn't seen that note before.
I know there are better solutions out there, but I'm in an
environment that thinks open source is the answer to everything.
<soap box>
By the time the project comes to me, the check has been written and
people are singing songs about low cost - high performance and
dancing in the streets.
Oracle's marketing seems to have embraced it too.
I'm sure you've seen the commercial where they imply you can replace
an expensive mainframe with a few "cheap" servers clustered together
with RAC...
when i'm in meetings and the BS siren goes off (seconds after the
salesman opens his mouth), I ask why they drive a Mercedes or
lexis' when I could put them in a new kia or ford? Their commercials
say they have the same features for a lot less.
a break is soon announced and I'm told they'll call me if they have
anymore questions.
I've come to the conclusion that these people (i.e. customers) still
believe in the Easter bunny.
there's another one I hear alot and that's "all databases are the
same - they're a commodity". That comes from managers with networking
backgrounds (now dba managers) who swear up and down we have to buy
Cisco. I asked why that same philosophy doesn't apply to networking.
the answer is "networking is totally different than databases"...
Engineering has always been a victim of sales and always will be. </soap box>
thanks again. steve
On Jul 23, 2006, at 04:02 PM, Kevin Closson wrote:
The only slogan I can come up with for this is "The land of the free and the brave". Folks, GFS forces you to have a node that does not host RAC instances dedicated to their lock manager.... and then that is a single point of failure...but I hear you can cluster the lock manager...is that 4 nodes for 2 node RAC?
All this central lock manager madness is just crazy. There are fully distributed, symetric, resilient, journalled CFS offerings out there. One runs on VMS and the other runs on Linux and Windows. I can tell you where to get the latter :-)
How "free" is that open source stuff afterall?
Reference Metalink 329530.1
I'm trying to avoid using ASM. It works, but I don't like having to use RMAN to do everything. There are more, but I'll let it go.
I've looked all over and can't find any references so I thought I'd check with the list. I was asked if we could use LVM2 (which is cluster aware) to manage disks in a RAC cluster. Instead of adding a new LUN and creating a new file system "/oraXX", they want to add a new lun to the volume group and extend it.
the RAC faq says it's not supported because lvm is not cluster aware, but that's LVM1 and RHEL3. this is LVM2 and RHEL4.
Does anybody know if it's supported? RH has notes on lvm2, gfs and RAC. I would hope they would publish docs on certified configs.
I was also considering OCFS2 and expanding the LUN (i.e. metalun), but can't find anything on how to extend the partition. I think they said after growing the LUN, use tuneocfs to extend the drive, but nothing about the partition. maybe I'm misreading it.
thanks, steve -- //www.freelists.org/webpage/oracle-l
-- //www.freelists.org/webpage/oracle-l
-- //www.freelists.org/webpage/oracle-l