Google+

Administrative Post and some Symmetrix V-Max discussions

April 26th, 2009 No comments

The past couple of weeks have been very busy for me related to work. Saw tons of announcements come out from EMC, VMware, Oracle, Sun, etc. My reading on those articles only made it to the headlines and a few paragraphs thereafter.
 

Today was a good day to catch up on all the blog readings. I have in excess to about 50 blogs I subscribe to. It has been very hard to read all the posts over the last two weeks. But today caught up on about 110 various blog posts from different storage / virtualization industry bloggers. 9 hours of nonstop reading.
 

Earlier this month, I changed my blogging platform from Blogger to Word Press. During the change, lost about 40% of the readers subscribed (feeds) to the blog.

To subscribe to StorageNerve Blog http://feeds2.feedburner.com/DataStorageProfessionals-Wiki

Read a bit about ORACLE-SUN, I am still confused about the entire bid, but Steve Duplessie did put a nice twist to the Sun purchase by Oracle, here…

Now to jump into V-Max……..read all blog posts from The Storage Anarchist, StorageZilla, Chuck Hollis, Chad Sakac, Powerlink and EMC V-Max page. Just amazing storage-techno stuff to read, I have committed to another 20 hours of faithful reading this week on the V-Max architecture and then write some blog post about it.
 

One thing I noticed with the V-Max blog post, every EMC blogger has been quite humble talking about the technology. On the other side, quite offensive stuff flying around from HDS and NetApp about the V-Max. Sorry I haven’t read any comments on any of the post, so can’t make a complete judgment.
 

I am sure you have read about the latest EMC announcement on V-Max, here is where you will find all the info…..
 

Technical Post / Articles

The Storage Anarchist

StorageZilla

Chad Sakac (VMware/EMC/Cisco)

Powerlink

EMC V-Max Landing Page

 

Technology

Chuck Hollis

Dave Graham

Some independent Bloggers have done a great job to overall summarize the Symmetrix V-Max announcement, here as referenced by Barry Burke.

 

Sometime in Feb 2009, I had written a blog post about expectations with EMC DMX next generation technology.

Did come quite close to predicting what the new set of architecture would look like, including Directors, Cache, IOPS, Drives types, Release dates, Cost offerings, GA enginuity code, Scalable Architectures, Controller Consolidation to include local memory, processor speed, EFD sizes, SMC console and associated Symapi, SMC friendless (templates, wizards), new conceptual design of mirror positions, virtual provisioning, power savings, Ethernet based connectivity to V-Max Service Processor, VMware Native Integration, Support for various RAID types that would enhance the product, policy based work flow automation, some support for FCOE in the future and a radical design change.
 

Things where I went wrong, the possible model numbers included DMX, Disk drive physical size (2.5 inch), PowerPC chip, no bin file, global memory, 8GB I/O interface, native support for Deduplication (I am sure you can stick a Celerra in front of a V-Max and achieve it).
 

Looking at the current GA Symmetrix V-Max product, I really did not expect the technology to go as far as EMC has managed to take it.
 

But that said a lot of hype has been around the V-Max architecture itself. One of the marketing buzz words is around scalability with hundreds of V-Max engines, 1000’s of TB’s of data, etc. At current GA, the product is scalable to 8 V-Max engines supporting 2 PB of storage and 2400 disk drives. EMC has been successful atleast in the first phase to sell the V-Max technology. With the current GA you are buying a future technology, not a technology that is currently in the product. Still not sure and haven’t been able to read on how the Virtual Matrix will connect the Physical V-Max Engines as they will span across geographies or even through larger datacenters.
 

Talking to various different customers over storage and storage technology, everyone feels pretty comfortable talking about HDS AMS, HDS USP-V, EMC Clariion, Compellent, Pillar Data and 3Par, but I still see customers struggling to talk about Symmetrix DMX, DMX-2, DMX-3 and the DMX-4 from a management and usability perspective. Will the V-Max change it all?
 

I would highly recommend a trip if you can to go to EMC World 2009. It will be great to see the V-Max in action along with the latest Celerra, CX4’s, Atmos, Cisco UCS and V-Sphere.

Transitioning from Blogger to WordPress

April 1st, 2009 No comments

Greetings from London this morning!!!

Over the past month and a half I had a chance to use WordPress with some work I have been doing with Gestalt IT, got really impressed with the interface, the ease of use and the end result in terms of the output.

Lately on Blogger Platform there have been a lot of issues with blog formatting, the posts layout and primarily the feeds. I have been traditionally writing post in MS Word and then copying it over to Blogger. It has something to do with the XML tags that get copied over along with the rest of the formatting, causing the feeds to stop delivering content to users.

Then the size of the feeds started kicking in, where the limit is 512kb. There was a blog post I did about blogger feeds and feedburner feeds with the size restrictions back in Jan 2009. It was a nightmare anywhere you turn around with that platform.

Tried do some work around WordPress couple of weeks ago, first of all impressed by how easy it was to import the posts from blogger. Within an hour all the posts and all the comments from Blogger were imported to WordPress. Started formatting the blog layout, it was easy. The output looks great, creating new blog post is easy, the feeds work and no real big issues. Currently I am running the built in WordPress version, if this works okay then over the next 3 months will migrate over to a hosted version that will allow running various different plug-in’s, etc.

But looking forward to working on WordPress. Hope the users find the content interesting, easy to read and feeds delivered to their RSS/ATOM readers.

Sometime during the day today on 04/01/2009, the domain will start pointing from Blogger to WordPress.

Storage Resource Analysis (SRA): Part 6

March 27th, 2009 No comments

TO SUBSCRIBE TO STORAGENERVE BLOG

Inconsistencies in Storage Environments

Continuing the blog posts on Storage Resource Analysis (SRA), this post focuses on some facts about what causes and what are inconsistencies in storage environments.
 

To read the previous blog posts on Storage Resource Analysis (SRA)

Storage Resource Analysis (SRA): Part 1: Storage Resource Analysis and Storage Economics

Storage Resource Analysis (SRA): Part 2: The IT – Storage World of 2009

Storage Resource Analysis (SRA): Part 3: The IT – Storage Budgets of 2009

Storage Resource Analysis (SRA): Part 4:Some Fundamental Questions

Storage Resource Analysis (SRA): Part 5:Facts about your Data

Storage Resource Analysis (SRA): Part 6: Inconsistencies in Storage Environments

Storage Resource Analysis (SRA): Part 7: The Technical Case

Storage Resource Analysis (SRA): Part 8: The Business Case

Storage Resource Analysis (SRA): Part 9: The End Result

 

To talk about a few inconsistencies that exist in the volatile storage environments, here is a subset of them, later in the post we will talk about what causes these inconsistencies.

o   Host masking to non existing LUNs
o   Host masking to invalid LUNs
o   Multipathing inconsistency with Hostmodes
o   Split BCV’s with no increments
o   BCV is smaller than Source devices
o   Source is smaller than BCV devices
o   Administratively fractured Clone copy, data integrity issues
o   Unallocated BCV to LUN
o   Host masked to LUN and LUN non mapped to path
o   LUN mapped to path and not masked to host
o   Single path host
o   Replication split
o   Replication failover
o   Replication Sync
o   R1 and R2 LUN to with improper host attachments
o   BCV never established
o   BCV Split
o   BCV Sync
o   BCV Mirroring
o   Empty disk drive slots
o   Disk drives installed but not used or configured
o   Physical Disk space (unused disk drive space)
o   LUN unallocated
o   Ungrouped disk
 

The above storage inconsistencies are pretty common with large environments and with multisite replication enabled. Also without a proper storage management tool, these errors are very likely to exist in any storage environment.

What causes these above set of issues; let’s talk about a few primary related reasons

o   Human error
o   Incorrect planning and implementation
o   Lack of Storage Strategy
o   Lack of Training
o   Lack of Reclamation Strategy
o   Storage Consolidation projects
o   Host migration projects
o   Host retirement or scrapped projects
o   Lack of Storage management
o   Operational Oversight
o   Undocumented planning and procedures
 

Experience

Just spoke to a potential customer last week. During a conference call we asked them, what are some of the major issues they are seeing in their storage environment? They have two Storage tools they use for operational and management purposes to handle a large double digit PB storage environment.

The answer from one of the architects was, we know of a lot of issues in our environment, but we have priorities around other things happening in the environment and cannot focus on these operational day to day non trivial issues.

With a large environment like the above to manage, a storage reclamation exercise can help a customer reclaim storage in terms of PB’s and could convert into immediate ROI, ROA and reduction in CapEx / OpEx that would help the organization save millions of dollars in new storage acquisitions. The question remains, are those our priorities today?

Do you have any of the above issues in your storage environment, or are you aware of them yet? 

Storage Resource Analysis (SRA): Part 5

March 27th, 2009 No comments

TO SUBSCRIBE TO STORAGENERVE BLOG

Facts about Storage


Continuing the blog posts on Storage Resource Analysis (SRA), this post focuses on some facts about data that sits in Storage Environments.

To read the previous blog posts on Storage Resource Analysis (SRA)

Storage Resource Analysis (SRA): Part 1: Storage Resource Analysis and Storage Economics

Storage Resource Analysis (SRA): Part 2: The IT – Storage World of 2009

Storage Resource Analysis (SRA): Part 3: The IT – Storage Budgets of 2009

Storage Resource Analysis (SRA): Part 4: Some Fundamental Questions

Storage Resource Analysis (SRA): Part 5: Facts about your Data

Storage Resource Analysis (SRA): Part 6: Inconsistencies in Storage Environments

Storage Resource Analysis (SRA): Part 7: The Technical Case

Storage Resource Analysis (SRA): Part 8: The Business Case

Storage Resource Analysis (SRA): Part 9: The End Result

 

Going to a very basic question about a Storage Environment, What percent of your storage is allocated and what percent of it is actually being used?

The more utilization you gain, the further you can push your new storage acquisition timeframes increasing your ROI and reducing your CapEx. The better efficiency and optimization you gain in your storage environments, the further you can push to reduce your OpEx.

So based on some latest data we have gathered from various different customer environments, believe it or not, on an average, this is what we saw…..

 

Allocated Storage

Average Customer Production data: Between 20 to 24%

Average Copy / Replication data: 20% more than Production data: Between 24 to 28%

Raid Overhead associated with production data: Between 15 to 19%

Storage Allocated but no owners: Between 18 to 22%

 

Unallocated Storage

Storage unallocated meant for future use: 13 to 17%

 

To shed some light on production and copy / replication data, there are a lot of data inconsistencies being reported in terms of broken links, replication failures, source and bcv lun size mismatches, etc.

 


 

The above scenarios with allocated / unallocated storage will reduce your ROI (Return on Investment) and ROA (Return on Asset), further causing budgeting issues in an organization.  

As you go to various different teams and ask them about the allocated / unallocated storage, you will hear different answers throughout the entire organization. From everyone’s view these numbers are debatable.  

 

Experience:

This time around, a personal experience.

In addition to my Technology Solutions responsibilities for our customers, I also have responsibilities around managing internal IT infrastructure that includes the apps, databases , storage and other IT assets. We have a lot of new projects happening every month, quarter, etc and have to allocate storage for those needs.

Projects that are hot today, might not be hot in the next 3 months, or may be in next 6 months. Projects get scrapped, host systems get scrapped, but storage still sits there, churning and spinning. Believe it or not, as the economic times change, lots of things are enforced and as the IT managers internally started looking at all the storage, we were able to reclaim quite a bit that was in real terms stranded storage, which is now all allocated to new ongoing projects or ready to be allocated at a short notice.

Do you know how much of your storage today is allocated, unallocated, stranded and importantly how much of it is Production data in your environment?