Comments for Frits Hoogland Weblog http://fritshoogland.wordpress.com IT Technology; Oracle, linux, TCP/IP and other stuff I find interesting Fri, 10 May 2013 12:52:32 +0000 hourly 1 http://wordpress.com/ Comment on When does an Oracle process know it’s on Exadata? by Latest data Industry news round up, Log Buffer #319 http://fritshoogland.wordpress.com/2013/05/04/when-does-an-oracle-process-know-its-on-exadata/#comment-2017 Fri, 10 May 2013 12:52:32 +0000 http://fritshoogland.wordpress.com/?p=947#comment-2017 […] When does an Oracle process know it’s on Exadata? […]

]]> Comment on Throttling IO with Linux by Kyle Hailey http://fritshoogland.wordpress.com/2012/12/15/throttling-io-with-linux/#comment-2015 Thu, 09 May 2013 23:24:35 +0000 http://fritshoogland.wordpress.com/?p=755#comment-2015 Awesome for performance bottleneck examples.
The linux distro I had 2.6.32 needed to have cgroup installed
yum install libcgroup

Thanks for posting this!

- Kyle Hailey

]]> Comment on Watching the “CopyBack” progress of a new disk on an Exadata compute node by Jee http://fritshoogland.wordpress.com/2013/05/03/watching-the-copyback-progress-of-a-new-disk-on-an-exadata-compute-node/#comment-2009 Sun, 05 May 2013 02:20:45 +0000 http://fritshoogland.wordpress.com/?p=950#comment-2009 Good Stuff. Also couple of reasons for not completing resync after systeem disk replaced, mentioned in 1279288.1 1328727.1 1316829.1.

]]> Comment on When does an Oracle process know it’s on Exadata? by Frits Hoogland http://fritshoogland.wordpress.com/2013/05/04/when-does-an-oracle-process-know-its-on-exadata/#comment-2008 Sat, 04 May 2013 21:29:14 +0000 http://fritshoogland.wordpress.com/?p=947#comment-2008 I ran a little test to look at the HCC verification. On a non-Exadata system, when I try to create a HCC table, the checking for a compatible tablespace is done in a function called ‘kdza_check_hcc_enabled0′. In that function kcfis_tablespace_is_on_sage is called to check if it’s on Exadata. The check is done on parse time.

This means that my current evidence does not suggest kcfis_tablespace_is_on_sage is used to probe for HCC capability if the query doesn’t involve HCC tables (because kdza_check_hcc_enabled0 is not called then).

]]> Comment on When does an Oracle process know it’s on Exadata? by Freek http://fritshoogland.wordpress.com/2013/05/04/when-does-an-oracle-process-know-its-on-exadata/#comment-2006 Sat, 04 May 2013 10:27:50 +0000 http://fritshoogland.wordpress.com/?p=947#comment-2006 Detection during parsing would make sense as the execution plan itself already shows if storage offloading is possible or not.

What does wonder me is why this is different on non exadata systems.
Does this mean that the database is maybe already checking if is is running on exadata on startup, and only verifies if the table (or tablespace) really resides on the storage cells (and not, for example, on nfs coupled storage)?
The kcfis_tablespace_is_on_sage check on non exadata systems could then maybe be explained as a check on the hcc capabilities of the storage (as HCC is now allowed on some non exadata storage systems as well).

]]> Comment on Watching the “CopyBack” progress of a new disk on an Exadata compute node by Freek http://fritshoogland.wordpress.com/2013/05/03/watching-the-copyback-progress-of-a-new-disk-on-an-exadata-compute-node/#comment-2002 Fri, 03 May 2013 16:52:03 +0000 http://fritshoogland.wordpress.com/?p=950#comment-2002 On X3-2 it seems that the compute nodes don’t have a hot spare anymore:

X3-2:
====

[root@XXXXXXX ~]# /opt/MegaRAID/MegaCli/MegaCli64 -pdlist -a0 | grep -iE “slot|firmware”
Slot Number: 0
Firmware state: Online, Spun Up
Device Firmware Level: A2B0
Slot Number: 1
Firmware state: Online, Spun Up
Device Firmware Level: A2B0
Slot Number: 2
Firmware state: Online, Spun Up
Device Firmware Level: A2B0
Slot Number: 3
Firmware state: Online, Spun Up
Device Firmware Level: A2B0

X2-2
====

[root@XXXXXXXXX ~]# /opt/MegaRAID/MegaCli/MegaCli64 -pdlist -a0 | grep -iE “slot|firmware”
Slot Number: 0
Firmware state: Online, Spun Up
Device Firmware Level: A2B0
Slot Number: 1
Firmware state: Online, Spun Up
Device Firmware Level: A2B0
Slot Number: 2
Firmware state: Online, Spun Up
Device Firmware Level: A2B0
Slot Number: 3
Firmware state: Hotspare, Spun down
Device Firmware Level: A2B0

]]> Comment on RMOUG and Hotsos by cstephens16 http://fritshoogland.wordpress.com/2013/04/05/rmoug-and-hotsos/#comment-1999 Tue, 30 Apr 2013 14:45:47 +0000 http://fritshoogland.wordpress.com/?p=920#comment-1999 Fantastic set of slides. Thanks for your work!

]]> Comment on RMOUG and Hotsos by Frits Hoogland http://fritshoogland.wordpress.com/2013/04/05/rmoug-and-hotsos/#comment-1993 Sun, 21 Apr 2013 07:51:22 +0000 http://fritshoogland.wordpress.com/?p=920#comment-1993 The Exadata OLTP paper is in the presentations and whitepapers section. Enjoy!

]]> Comment on Using LVM snapshots for backup and restore of filesystems by Frits Hoogland http://fritshoogland.wordpress.com/2012/08/11/using-lvm-snapshots-for-backup-and-restore-of-filesystems/#comment-1989 Tue, 16 Apr 2013 15:06:07 +0000 http://fritshoogland.wordpress.com/?p=684#comment-1989 Joe, your writing style is a bit bullying, are you aware of that?

I did test this method, but the lack of details in your reply makes it hard to reply.

A summary of this post is:
A snapshot contains a pre-upgrade state.
If the upgrade fails, the lv and snapshot differ, because the lv is post-upgrade.
I am aware what a snapshot is, but it’s about the state of the filesystem in the snapshot.
The current lv is renamed, and the snapshot contents are copied from the snapshot to a new lv.

A potential problem is the snapshot contains a mounted filesystem.

If you are able to produce a reply with technical content (beyond ‘doesn’t work’) I am happy to discuss this with you.

]]> Comment on Using LVM snapshots for backup and restore of filesystems by joe http://fritshoogland.wordpress.com/2012/08/11/using-lvm-snapshots-for-backup-and-restore-of-filesystems/#comment-1988 Tue, 16 Apr 2013 13:45:08 +0000 http://fritshoogland.wordpress.com/?p=684#comment-1988 Well, dd’ing an LVM snapshot volume is useless because the snapshot volume only contains the COW copies from changes to the original LV! (If there were no changes to the original root LVM, then the snapshot would essentially be empty. So you want me to copy a COW repository, which potentially contains only possible changes back as an original LV? What are you thinking? I tried this experiment on a few servers and as expected, fails miserably. I believe this is one of those situations where somebody just publishes b.s. without really testing it or knowing what they are talking about.
Cheers

]]>