Comments on: Little Things Doth Crabby Make Part I. Enterprise Linux 5/RHEL5 Output Format Change for the iostat Command http://kevinclosson.wordpress.com/2008/03/08/little-things-doth-crabby-make-enterprise-linux-5rhel5-output-format-change-for-the-iostat-command/ Oracle-related Platform, Storage and Clustering Topics (with the occasional rant) Sat, 17 May 2008 23:53:59 +0000 http://wordpress.org/?v=MU By: Little Things Doth Crabby Make Part III. Non-Erroring Errors and Erroneous Experiments. « Kevin Closson’s Oracle Blog: Platform, Storage & Clustering Topics Related to Oracle Databases http://kevinclosson.wordpress.com/2008/03/08/little-things-doth-crabby-make-enterprise-linux-5rhel5-output-format-change-for-the-iostat-command/#comment-32573 Little Things Doth Crabby Make Part III. Non-Erroring Errors and Erroneous Experiments. « Kevin Closson’s Oracle Blog: Platform, Storage & Clustering Topics Related to Oracle Databases Fri, 28 Mar 2008 19:31:55 +0000 http://kevinclosson.wordpress.com/?p=545#comment-32573 [...] post, but its use is not the central theme. No, no Silver Bullets here. This is another post in the Little Things Doth Crabby Make [...] [...] post, but its use is not the central theme. No, no Silver Bullets here. This is another post in the Little Things Doth Crabby Make [...]

]]> By: kevinclosson http://kevinclosson.wordpress.com/2008/03/08/little-things-doth-crabby-make-enterprise-linux-5rhel5-output-format-change-for-the-iostat-command/#comment-32542 kevinclosson Sat, 22 Mar 2008 05:08:41 +0000 http://kevinclosson.wordpress.com/?p=545#comment-32542 Opening a raw device (raw(8)) with open is the same thing as opening a block device with open(O_DIRECT). There is no "there" there. ...thanks for stopping by Alex...as always. Opening a raw device (raw(8)) with open is the same thing as opening a block device with open(O_DIRECT). There is no “there” there.

…thanks for stopping by Alex…as always.

]]> By: Alex Gorbachev http://kevinclosson.wordpress.com/2008/03/08/little-things-doth-crabby-make-enterprise-linux-5rhel5-output-format-change-for-the-iostat-command/#comment-32538 Alex Gorbachev Sat, 22 Mar 2008 01:35:38 +0000 http://kevinclosson.wordpress.com/?p=545#comment-32538 Luke, I wish all SA's would be more like you. Not many SA's are reading Oracle blogs. :) Re: raw devices. With 10gR2 you need to configure raw devices only during installation of 10.2.0.1 CRS - voting disks and OCR. After upgrade to 10.2.0.3, I always get rid of raw devices and use block devices directly. They are opened with O_DIRECT. I like to exclude unnecessary layers and raw devices is a good example. Luke, I wish all SA’s would be more like you.
Not many SA’s are reading Oracle blogs. :)

Re: raw devices. With 10gR2 you need to configure raw devices only during installation of 10.2.0.1 CRS - voting disks and OCR. After upgrade to 10.2.0.3, I always get rid of raw devices and use block devices directly. They are opened with O_DIRECT. I like to exclude unnecessary layers and raw devices is a good example.

]]> By: Luke Youngblood http://kevinclosson.wordpress.com/2008/03/08/little-things-doth-crabby-make-enterprise-linux-5rhel5-output-format-change-for-the-iostat-command/#comment-32508 Luke Youngblood Sat, 15 Mar 2008 00:10:43 +0000 http://kevinclosson.wordpress.com/?p=545#comment-32508 Your blog is great and very interesting. Here is my experience. I'm a Linux sysadmin (not a DBA) but all I do is Oracle RAC support for our DBA team. We've used Redhat 5 and 11g on a single test server, but our head DBA says we're holding off on it because Oracle hasn't yet certified Grid Agent as compatible with Redhat 5. I'm deploying 3 brand new 4-node clusters on Sun Fire X4600 (8 way Opteron baby!) with EVA 8100 storage on the back end, quad 4gb HBAs in every server, and we are going Redhat 4 Update 6 64-bit and Oracle 10gR2 all the way. Also, I have to say it's a major pain in the ass how they deprecated the /etc/sysconfig/rawdevices interface. It was easy to configure before and now it's just a pain in the ass. Well, I'm sure Redhat 5 will be nice eventually, but for now, all we get is: 1. A scheduler in the new kernel that is not very database friendly (Completely Fair Scheduler is nice for desktop Linux but terrible for database). 2. Built in virtualization (woohoo! Now my single purpose Oracle box has the capability of running more than one OS, which it will never do...) 3. A deprecated rawdevices interface that makes life difficult for Oracle DBAs and sysadmins. 4. And as you mentioned, all of my performance monitoring scripts that pull data out of iostat and INSERT it into a table have to be modified. [edited] Keep up the blog, this is great stuff. Your blog is great and very interesting. Here is my experience. I’m a Linux sysadmin (not a DBA) but all I do is Oracle RAC support for our DBA team.

We’ve used Redhat 5 and 11g on a single test server, but our head DBA says we’re holding off on it because Oracle hasn’t yet certified Grid Agent as compatible with Redhat 5.

I’m deploying 3 brand new 4-node clusters on Sun Fire X4600 (8 way Opteron baby!) with EVA 8100 storage on the back end, quad 4gb HBAs in every server, and we are going Redhat 4 Update 6 64-bit and Oracle 10gR2 all the way.

Also, I have to say it’s a major pain in the ass how they deprecated the /etc/sysconfig/rawdevices interface. It was easy to configure before and now it’s just a pain in the ass.

Well, I’m sure Redhat 5 will be nice eventually, but for now, all we get is:

1. A scheduler in the new kernel that is not very database friendly (Completely Fair Scheduler is nice for desktop Linux but terrible for database).

2. Built in virtualization (woohoo! Now my single purpose Oracle box has the capability of running more than one OS, which it will never do… ;)

3. A deprecated rawdevices interface that makes life difficult for Oracle DBAs and sysadmins.

4. And as you mentioned, all of my performance monitoring scripts that pull data out of iostat and INSERT it into a table have to be modified.

[edited]

Keep up the blog, this is great stuff.

]]>