This year at EMC World, I was asked along with one of my BRS Integration Labs colleagues to produce a Lab for the Hands on Labs that walks a user through using Oracle with dNFS and Data Domain BOOST, and reporting it all with Data Protection Advisor (DPA). I am publishing the entire lab guide here
As always please leave feedback with questions.
I am posting my EMC World presentation for the session Caitlin Gordon and I presented this week. Feel free to ask questions or provide feedback here.
Get the powerpoint from my Syncplicity (shameless plug) here
So as I begin my journey into learning about Openstack, i saw recently that RedHat has a Openstack distro. So I deployed a CentOS 6.4 VM in my home lab. I went to the Redhat page here.
I followed the instructions but kept running into this error.
ERROR : Error during puppet run : err: /Stage[main]//Exec[setenforce 0]/returns: change from notrun to 0 failed: setenforce 0 returned 1 instead of one of [0] at /var/tmp/packstack/23da28e64ec9446aac7c440dccc49b89/manifests/192.168.0.90_ring_swift.pp:56
The keypiece of the error is this setenforce 0
In my CentOS 6.4 VM template I have SELinux disabled. Apparently this error is not liking that. So you have to enable SELinux by doing the following.
edit the /etc/selinux/config file and change the following paramater
SELINUX=enforcing
then you have to REBOOT to make the changes effective.
I was then able to run the
packstack --allinone
and it completed successfully.
This is a shameless plug for my EMC World session.
I have been asked to present with Caitlin Gordon on Data Domain and Oracle.
We have worked very hard on the content, I even have some VMware related content I was able to sneak in. Here is the session info you can find on the emcworld.com website.
Caitlin and I have worked really hard on trying to make this a very technical session, with very little marketing fluff. If you know me, its that I always want the attendee to walk away with knowledge they can take back to their own environment and make use of.
As of today here is the date/time and location.
Monday 11:30AM Palazzo D
Tuesday 10:00AM Palazzo M
Hope to see you there!
After much debate I have decided to post the presentation that I used to give to the vSpecialist and VCE vArchitect newhire classes. This is a little dated, as it is for the original VMAX now called the 20K, and Enginuity 5875 code. I have scrubbed anything that might possibly be NDA. This info is around in various places but I thought it would be nice to post it for the world in an effort to be more transparent about how a VMAX works. Please send me any feedback you have around it.
UPDATE
@codyhosterman has advised me there are a few corrections to be made, I will update this post when I have made the edits to the PPTX. Thanks Cody!
Lately I have been working a lot with building a Oracle database template for our lab. And have been working to optimize IO inside of these guests. So i found a few things I wanted to pass on.
First thing is partition alignment. As of RHEL6 the partitions are correctly aligned at 2048 sectors. Which is a great thing!
But make sure when creating new partitions with fdisk to change your Units using the u command and then specifying the starting offset at a multiple of 64 like 2048.
The second is using different IO Schedulers or elevators. This used to be achieved by adding a deadline= parameter at the end of the kernel loader statement in GRUB. But now there is a new way by using a command called tuned-adm
here is great explanation of what it does. Tuned and ktune
and also how to use it here
I had been using the enterprise-storage profile up until 6.3. Now i am using the virtual-guest which is based on the enterprise-storage but updated as the article explains.
I will update this post in the future when I get around to do doing some real-world performance testing using Swingbench.
Just wanted to drop a quick note that as of today the a new Service Release (SR) Of Enginuity microcode has been released which includes support for the vSphere VAAI primitive UNMAP. This has been a long time coming See this blog post here explaining why the feature was pulled.
One significant note, the UNMAP operation does not happen automagically. You have to use a command to manually invoke it. See this VMware KB
This Enginuity release also introduces other new features like an updated Unisphere for VMAX 1.5. You can find the detailed release notes on Powerlink here
Home > Support > Technical Documentation and Advisories > Hardware/Platforms Documentation > Symmetrix VMAX 20K/VMAX Series > Release Notes
And wait! There is more!
There is also a refreshed VMAX 10k hardware included with this release. More cores in the Intel CPU, and the ability to mix and match 2.5″ and 3.5″ drives in the same disk array enclosure (DAE).
Federated Tiered storage, Data at rest encryption and more granular IO control are also included.
So a very feature packed release that you should look into if you are using a VMAX.
Just in case anyone else was wondering, after the Synology issue where they apparently did not test it with vSphere 5.1.
I am indeed running Nexenta version 3.1.3
And vSphere 5.1 GA build. I have not run into any issues, all of my VMs are running fine. I am using iSCSI by the way.
So I am upgrading my home servers via VUM, and i noticed a screen on the console after the upgrade finished, it stated that the host must reboot. and to hit Enter to continue. There was no timer on the screen, so after waiting approx. 5 minutes the server rebooted on its on. It would have been nice for there to be a timer on the screen so the user would know the system would eventually reboot on its on. Just my 2 cents.
So I have gotten this question quite a bit since I announced my move from the vSpecialists. So I thought I would try and explain a little bit about my new team.
The BRS Integration lab, is located in Hopkinton, MA across the street from the corporate office. It is a lab (obviously) that has EMC arrays (duh) but also many competitors and more servers than you can shake a stick at. Its main purpose is to provide the EMC field with a place to take a customer request, say backing up 1000 VMware VM’s on a Avamar grid, from a VNX 7500, and prove it works. So by now you I am sure you are asking “but you are a VMAX and VMware guy, what does this have to do with any of that?”
Good question! I am not going to the BRS Integration lab because of my mad backup skills (I do have experience in backup). But more to assist with the VMware integration involving backup. Things like, “how do we protect VMware View, or vCloud Director, or vCloud Data Director, etc.”
And of course we have 2 VMAX’s. So with this new role i will be working more on VMAX and VMware than I ever have in my career. Also the lab helps with development of new products and solutions. I hope this gives everyone a little more insight as to what we will be up to.
You can be sure I will be posting things as I am able to (lots of NDA stuff).