I worked with a vendor to install a new HP c7000 enclosure and FlexConnect switches. As part of the preparation and as my homework, I read through the HP Virtual Connect FlexFabric Cookbook to become familiar with this technology. If you are thinking about moving to Virtual Connect, you should be familiar with this document. There are many others you should read as well. All of them are available on HP’s site but are not very easy to find. It seems Google or Bing make it easier to find than searching from HP’s site.
I read through the FlexFabric Cookbook, Virtual Connect Ethernet Cookbook, Virtual Connect for c Class, and Virtual Connect iSCSI Cookbook. After this, I thought I had a base understanding of the technology from reading and prior experience with c Class Blade System and blade switches. I am a Cisco guy and used to working with their IOS based switches and modules. I thought I had a good understanding of how this would work but I could not have been more wrong. After struggling with the vendor and HP Support to resolve issues, I received help via HP Support’s Twitter account (@HPSupport). They were able to get me in touch with the a wonderful engineer who took all the time in the world to explain FlexFabric in a way I would understand. As a side note, I have found that the support or service Twitter account to be an invaluable resource for assistance.
This engineer also put me on the a hidden piece of treasure, HP Virtual Connect for the Cisco Network Administrator. This was the missing piece of the puzzle! Again a search on Google turns this document up quicker than a search on HP.com.
Armed with the information from the engineer and the document, I am now successfully navigating the FlexConnect world.
If you have difficulty finding any of the documents mentioned in this post, ping me and I will be happy to direct you toward them.
Lately I have seen a rash of error messages from my VNX systems reporting warning level messages on drives with the description “Soft Media Error”. After searching EMC Support site, I could not find a good link with my error. I searched the Community site and still could not turn up an entry with a solution.
I would like to say that the best solution in any case in which you cannot find a solution is to open a service request. The support technicians will get to the bottom of the issue quickly but may not always offer a lot of information for future occurrences.
In my case, these were predictive failure warnings. This could be seen by analyzing the collects from the system.
After opening a service request and receiving the replacement disk, I could not find a lot of information with how to replace a disk that had not actually failed. I present what I was able to pull together on proactive disk replacement.
In order to perform any maintenance on a VNX, you will need to have access to Unisphere Service Manager which is available to download from the EMC Support site downloads.
NOTE: USM will provide options to use LEDs to identify the array and disk with the error.
I received warnings in Microsoft System Center Advisor (a proactive monitoring tool included with Microsoft Premier Support) about my databases running in SQL Express. How do you manage and maintain those SQL Express databases?
First, remember that although it is not the full version of Microsoft SQL, it does still require patching and maintenance. Windows Updates should show SQL service packs and patches in available offerings. As a default plan, I try to keep my SQL Express databases on the same patch level as my full SQL Servers. (Unless other limitations apply). If you have access to Microsoft System Center Advisor, it will keep you updated for any SQL Express installations on Windows Server 2008 R2 or newer.
Microsoft provides a version of SQL Server Management Studio for SQL Express in Microsoft Downloads. It is listed as Microsoft SQL Server 2008 R2 Management Studio Express. Although SQL Server Express does not allow for Maintenance Plans, you can still manually run scripts for maintenance tasks such as checkdb, reindex, backup, etc.
After installing Management Studio and gaining access to the database I could manually backup the database and run some management tasks with scripts I created earlier. To maintain the health of the MSONLINE database, I wanted to backup the database, check the consistency, and reindex the database.
To accomplish this, create a SQL script to backup all the databases on your instance, one to run checkdb, and one to reindex. (You can search the internet to find examples or script database in Management Studio). Save these scripts in a folder locally on the server. In my example, I use C:\SQL_Scripts\MSONLINE.
Create a Windows Command Script file to run the SQL scripts. This will allow you to run this as a Scheduled Task in Windows Server. In my script, I add all three maintenance tasks as shown below and direct output to a report file in the same folder:
REM Backup databases
sqlcmd -E -S servername\MSONLINE -i”C:\SQL_Scripts\MSONLINE\BackupDB.sql” -o “C:\SQL_Scripts\MSONLINE\Backup.rpt”
REM Check Database
sqlcmd -E -S servername\MSONLINE -i”C:\SQL_Scripts\MSONLINE\CheckDB.sql” -o “C:\SQL_Scripts\MSONLINE\CheckDB.rpt”
REM Rebuild Index
sqlcmd -E -S servername\MSONLINE -i”C:\SQL_Scripts\MSONLINE\Reindex.sql” -o “C:\SQL_Scripts\MSONLINE\Reindex.rpt”
You may also want to add a script to automatically cleanup your backup folder.
This should help you keep your Directory Sync SQL Express database running and healthy.
Check out Microsoft System Center Advisor at https://www.systemcenteradvisor.com/. For more information on Microsoft Premier Support see http://www.microsoft.com/en-us/microsoftservices/support.aspx
I ran into an issue recently with a corrupted installation of Microsoft Directory Synchronization tool for Office 365.
In the past, support would tell us to try running the configuration wizard again and if that failed uninstall the tool and reinstall.
I have multiple databases on my instance of SQL Express.
To remove, run Uninstall a Program.
Select Microsoft SQL Server 2008 R2 (64-bit).
Select the instance to remove. In this case MSONLINE.
Select Features to remove. If it is your only instance, click the Select All button. If you have other instances, select only the features under your database, not the Shared features.
This will remove the instance installed by the Directory Synchronization tool, MSONLINE.
You can then run Directory Synchronization tool installation again and configure.
Here are the Facebook pet peeves of a grumpy old man:
So there you are. Now we can all be friends again. Well, at least Facebook friends.
My last post in the Blogger Reality Show 2011 prompted several comments and questions about old fashioned molasses stack cake. It sparked old, tasty memories for some and questions and Google searches for others.
It was one of my favorites growing up possibly because it was a rare treat. I never understood why it was a special cake until I read the recipe. Seeing the time it takes to make it (and a proper one should have no less than 10 layers) I understand why this was not the cake of choice for Sunday dinner each week! Several regions lay claim to its origins. I will leave that argument for others to settle. I will simply say that all of my Appalachian friends have heard of and had this cake.
Rather than type the recipe for you here, I give you the recipe from my mother’s own hand from one of the man yellowed spiral notebooks full of Appalachian cooking at its best. Remember to invite friends over when you make it.
Although I can’t indulge as I would like to, I still love desserts. I love all kinds but nothing beats a good piece of cake. My great grandmother, grandmother, and mother all used to make an old fashioned molasses stack cake. Instead of two or three thick layers, this cake is made up of multiple thin layers of cake. Virtualization in the data center reminds me of this cake. What were once two or three layers in the data center are now multiple layers many of which are virtualized.
Server virtualization presents the processor, memory, network card, and CD/DVD drives of a host as virtualized components to many virtual machines. Some have said storing data in RAID sets represents data virtualization at the lowest level by storing it simultaneously on multiple physical disks. RAID sets can be duplicated or replicated across groups of disks in the same storage device or scaled out to other storage devices. In the most recent webcast in our Blogger Reality Show Contest, we take a look at the remaining piece of this virtualization puzzle, network virtualization. Lastly, we take a look at how we manage these pieces.
Virtual Connect technology provides access to the LAN and SAN through flexible fabric in the blade system. It simplifies hardware needed such as switches and cables. It also simplifies changes. The “wire-once” technology simplifies processes through server profiles. In essence it virtualizes the networking layer of the server, layers 1 and 2 in the OSI model. Other vendors such as Cisco with UCS, and VMware Server Profiles offer solutions in this space as well. HP approached this from a server admin point of view. Server Profiles allow you to take the MAC address, storage information, boot, PXE boot, and serial numbers from a physical blade server and apply this information to a device bay in the blade chassis thus virtualizing the information. It is no longer tied to the physical blade. This allows you to replace the physical blade without reconfiguring all of this information. This information can be shared between c7000 blade enclosures in a stack or across stacks with Virtual Connect Enterprise Manager. It supports up to four c7000s in a stack (or Virtual Connect domain). You can see the power of this for rapid deployment of blade servers, managing workloads, and failing over an entire chassis. HP’s solution also provides in rack optimization by containing most traffic within the chassis or between chassis thus reducing traffic to top of rack switches. Virtual Connect does prevent loops by using active loop prevention technology similar to NIC teaming rather than STP.
The FlexFabric Modules and Virtual Connect Manager answer the question of how this is accomplished. From the Flex Fabric Module converging Ethernet, Fibre Channel, or iSCSI to the Flex-10 Module converging Ethernet and iSCIS or providing separate Ethernet and Fiber Channel fabric the modules allow multiple networks across blade hosts on a single device with minimized cabling.
The Flex-10 Module has provided a leap forward in virtualization by allowing all of the networks needed to be converged on one device with one cable. A single 10 GB connection can be divided to provide a management network for service console access, a network for VMotion, a storage network, and one for the virtual machines. In addition to providing the FlexNICs for these networks, it provides the HBAs for the storage network.
Both the standard manager and enterprise manager display the common HP look and feel. These managers are where you will configure the modules and Virtual Connect domains. The enterprise manager centralizes management for multiple c7000 chassis. Some management features are also available by Insight Software add-in for vCenter.
Years ago, I had a table beside my recliner just for all of my remotes. I had a remote for the TV, one for the cable box, one for the DVD, and one for the sound system. I thought all of my problems were solved when the universal remote came out. I was just as quickly disappointed. Anything “universal” can do many things but none of them well. At best I was able to reduce my remote collection by two. The universal remote was able to power on and off all of the devices and handle basic functions such as changing channels, adjusting volume, and playing a DVD. However, I missed many of the advanced features on each that I had come to rely on.
As many of you are, I find myself overwhelmed with management software and interfaces. Each vendor’s management interface has a completely different look and feel. Most of us have longed for a true universal management system but it has never been delivered. There are many that are close but like the universal remote, they lack many features we rely on. Most can give you up/down status and some basic information through pinging, SNMP collections and Syslogs. You can get some network information through NetFlow. I have used SolarWinds Network Performance Manager and many of their modules in addition to Microsoft System Center Configuration Manager and Operations Manager. None have given me the features I would like and I end up going back to the various vendor interfaces. Another shortcoming of most is that they are Java based. I don’t need to repeat the frustrations I am certain all have experienced with Java based apps and different Java versions.
With IMC, HP is delivering a solution to monitor, manage, and provision the network; a single plane of management. It is a modular design with additional add-on features that will work with many vendors management interfaces. It provides reporting and analysis of data gathered. Through user configuration, it provides a unique view for each user customized to that user’s function or interest. It is a Java based tool but did not seem to suffer through many of the slow downs during the demonstration that I have experienced in my environment. HP is building this to meet future standards and needs of the virtualized data center as well.
I hope HP can succeed where others have come up short.
These final pieces of network virtualization and management systems complete the puzzle of the new data center which what I consider the cornerstone of a cloud deployment. HP brings their unique approach to both. HP is one of few technology companies positioned to provide solutions in each area of virtualization in the data center from servers, storage, networking, to management.
As a long time enterprise HP customer, I have a lot of familiarity and experience with many of their products. This contest offered me the opportunity to learn about products that may not currently be in my environment or on my roadmap. I have gained a greater knowledge of current technology and trends and it has been my pleasure to share with you. I would like to thank Thomas Jones (aka NikeTown588), HP, and Ivy Worldwide for putting together this contest. I would also like to thank the other bloggers competing and the judges for all they have done to bring information and value to this community of technology leaders, users, and enthusiasts.
Please take a moment to vote in the poll below and show your support. Thank you.
This post is the third and final post in a series as part of Thomas Jones (aka Niketown588) “Blogger Reality Contest”. The contest is sponsored by HP and Ivy Worldwide. As a contestent I will receive an expense paid trip and conference pass to VMworld 2011.
At the risk of appearing to be uncool and not in the know, I want to ask the question what is cloud?Is “cloud” a tangible product? Is it a service provided by a vendor? Is cloud a way of designing your architecture? Is it a marketing term? Is it a destination? Is cloud an IT management philosophy? The answer depends on who is answering the question.
In a recent blog post from Chuck Hollis, What’s In A Name, he takes a look at information technology as information management and as “maximizing the value of their corporate information portfolio. The technology used to do this was simply a means to an end, and not the end unto itself”.
I see this transformation of IT to IM in many SMB companies I am familiar with as well as my own. It has caused me to look at the infrastructure we have in place today and the plan for where we want it to be in a different way. Is it the means to an end? Issues such as a greatly dispersed workforce, a diverse business portfolio with equally diverse business needs, and information silos which must remain in place and be managed all point to a need for a different way of looking at the underlying technology. The means to deliver the solution to these needs is our definition of cloud.
I work for a company in the SMB range and as most companies operate these days, we try to do more with less. An important aspect to any solution I look at is how easily the solution goes from a crate to production. How much time does the solution take to manage on a day to day basis?
Some of the most popular solutions out there come in the form of stacks from vendors such as Dell, IBM, HP, Oracle or coalitions such as VCE, These solutions are designed to be a controlled, turn-key solution that some refer to as “cloud in a box”. It is a preconfigured and certified solution containing storage, servers, networking, and software designed to meet your system or application needs. It fits the needs I have today by minimizing configuration and management.
The webinars we have viewed so far in this Blogger Reality Show Contest from HP have given us a look at HP’s approach to the infrastructure and platform for cloud and this new way of thinking and new way of delivering technology. In week one, we looked at Converged Storage (my blog post).
HP attempts to move beyond the silos within a datacenter by providing SAN/IQ which can bring together different storage products and platforms. With their converged systems, HP has designed their solutions to break out of the stack by making their solutions flexible enough to run any of the most popular hypervisors from Citrix, Microsoft, and VMware. Beyond that point, it has many
similarities to the other cloud in a box systems; Blade system for compute, storage solution supporting virtualization, and interconnect-type networking system for “in the box” connectivity.
HP’s portfolio of converged systems includes HP VirtualSystem, HP CloudSystem, and HP AppSystem. Each are built on solid HP components, c7000 Blade System, P4000 Storage Systems, and Virtual Connect multi fabric system. Each is managed with HP’s Insight Control Environment management system. If you are familiar with any of HP’s management systems, you will find it has a very familiar look and feel. However each is designed to fit a different need.
This solution fits in my world of small to medium businesses. It is a multipurpose, modular system designed to run 750, 2,500, or 6,000 virtual servers or approximately 3,000 virtual clients. This solution would be ideal to host virtual servers, applications, or virtual desktops in the datacenter.
The CloudSystem is designed to get a private, hybrid, or hosted public cloud up and running quickly. All flavors contain the underlying Blade System and Matrix Operating Environment. They differ as they grow including Cloud Service Automation, and Aggregation Platform for SaaS network automation. CloudSystem comes in three configurations to meet different needs. They are CloudSystem
Matrix, CloudSystem Enterprise, and CloudSystem Service Provider versions. My expertise is limited to smaller environments so I focused on CloudSystem Matrix. HP again is speaking to my needs by emphasizing making it easy to quickly provision this system. For those with a tighter schedule, HP offers CloudStart. It is service offering to get you up and running within 30 days of installation. CloudSystem Enterprise allows you to scale out to public cloud. CloudSystem Service Provider, like its name indicates, is designed for the service provider.
The final solution we look at is the AppSystem portfolio which is designed to fit needs in data warehousing, data analytics, and business intelligence. When looking at solutions to fit these needs, one key component is storage capacity. As you warehouse or analyze data it obviously requires an increasing amount of space.
HP designed these solutions to scale out to additional units or scale out, or burst out, to public cloud space.
The systems are designed and certified to work with Microsoft SQL family and HP’s on analytics system, Vertica. If you’ve looked at solutions in this space you will see a trend in providing systems that are simpler to get up and running and very scalable. This is a theme near and dear to my heart. HP hits this them by simplifying management, scaling, and licensing. The Microsoft solutions are unlicensed as many of us already have Microsoft Volume Licensing agreements in place.
However you define “cloud”, you will see changes in the way information technology is viewed within our businesses, changes in expectations about the use of information, and changes in the technology used to deliver information. HP and other technology leaders are designing solutions to get us there.
This post is the second in a series as part of Thomas Jones (aka Niketown588) “Blogger Reality Contest”. The contest is sponsored by HP and Ivy Worldwide. As a contestent I will receive an expense paid trip and conference pass to VMworld 2011.
There are many buzzwords and phrases flying around technology today. Cloud seems to be one of the mosed used, misued, and abused. A close second has to be ____ as a Service. The latter has mostly become IT as a Service. Once you look at information technology as a service we move our thinking toward how we provide access to needed information and less time spent on the traditional structures that we used to spend out time on. Those structure become a commodity. There a many types of these structures you can use to provide services. You must evaluate them to find the ones that fit your business needs and budget.
Server virtualization began our journey which then continued with changes to networking. This has resulted in making physical location of servers irrelevent to the consumers of the service. One of the biggest remaining elements in this journey to virtualization of structures is storage. It is hard to overcome that although the services can be located anywhere and accessed by consumers from anywhere with multiple decvices, the fact remains that theses services are still actually phsically located on disks somewhere. Security, availability, and reliability are critical at this level. This level is now seeing innovation and making leaps forward after not changing much over the past years. The explosive growth of data is also fueling this need for innovation.
To overcome these silos of storage, we first address silos within a datacenter and then silos by location. In this webinar, we looked at HP’s approach. HP is positioned to provide products and servcies in each area of these covergerd infrastructures with their server, networking, and storage offerings. They are building systems based on industry standard x-86 architecture using their c-7000 class blade system. These systems more easily scale up and out without deadending you on a platform which limits growth and forcing costly replacements in the future. HP uses a software based solution, Storage Provisioning Manager and Storage Essentials Manager, to enable storage management accross platforms and vendors. These also plug in to other managment systems such as vCenter. HP’s software uses a common layout so if you are familiar with any of their software such as Systems Insight Manager, you will find your way around.
In this webinar we took a look at the HP P4800 G2. As stated above, this system is built on the c7000 Blade Enclosure . This provides advantages of shared resources improving efficiencies in power and cooling. Converged networking through multi fabric switching options are one of the primary advatages. The interconnects built in to the system allow you a lot of options such as HPs Flex 10 technology as described in this technology brief. The guts of the system are the MDS-600 storage system. It is a SAS disk based blade system storage system.
The brains of the system is the SAN/iQ software. This brings together storage from different arrays into a converged storage system. SAN/iQ also handles the copying/replicating of the data between sites. Based on distance, bandwidth, and thus latency this can be synchronous, asynchronous, or an off site replictation.
The P4800 shows how far we have come in the journey to virtualized systems. HP has built solutions to break through local storage silos and move toward scaling out to other sites. We also still have further to go. We still have limitations on storage making it challenging to make storage transparent. In my position, I fight storage silos popping up and many remote sites. The solution I need will be the one that can address this.
This post is the first in a series as part of Thomas Jones (aka Niketown588) “Blogger Reality Contest”. The contest is sponsored by HP and Ivy Worldwide. As a contestent I will receive an expense paid trip and conference pass to VMworld 2011.
When looking for a snarky comment or to see what the cool kids would say, I always refer to Despair.com which offers a take off on the once popular Covey motivational posters. Their take on blogging: “Never before have so many people with so little to say said so much to so few”. While that can be true for some, I have learned to find what church technology visionary Anthony Coppedge (found here) “signal in the noise”.
Anthony made this comment about using Twitter effectively. Twitter was orignially described as micro-blogging. Twitter seems to have eclipsed blogs in both popularity and relevance. I think of blogging now as macro-Twitter. Blogging allows us a longer format to continue the conversation. If you are in technology or interested in technology, hopefully by now you have found several influential people to follow on Twitter and blogs to read. These formats offer a means of communication that can finally keep up with the pace of technology. It is much better than keeping tech newsletters and magazines around and finding time to read them.
This brings me to the purpose of this post. A friend, technology blogger, podcaster, and Twitter leader, Thomas Jones (of NikeTown588 fame) has come up with a contest sponsored by Hewlett Packard and Ivy Worldwide. It is being called a blogging reality show except none of us will be banished or forced to live in a house with strangers. So far seven of us have been chosen to particpate. You can read further about the contest in Thomas’s blog post here. The basic rules are we are to view three instructor lead classes on three different subjects and write about it on our blogs. You the readers get to vote. Your votes along with votes from a panel of judges will be added up and winners will be announced at VMworld 2011 in Las Vegas the week of August 29 through September 1.
I entered this contest with a desire to sharpen my blogging skills especially in reference to technology. I am also excited for the opportunity to learn more about technology and have the opportunity to ask questions of experts and share with you what I have learned. You, the reader will gain from this by reading not only my blog posts but those of the other competitors. I encourage you to read all the blogs in this contest not just to vote but to learn about the subjects we are covering.
We are all part of a community of technology users, leaders, analysts, experts, etc. We participate in the community by reading what others have written, sharing, commenting, or writing things ourselves. In order to maintain a thriving community we need to make sure we reach out to find people who are interested. John Maxwell said “If it’s lonlely at the top, you’re not doing something right. Who did you bring with you?” If you listen to Thomas’s pocast Coffee With Thomas, you know he has a following of young people and women in technology. This is an important audience to him. He is looking for interested women to participate in this contest. I would encourage anyone interest to contact him and join us. Lets make an effort to include everyone in our efforts.
Again, I would like to thank Thomas, HP, Ivy Worldwide, the other participants and you readers. I look forward to sharing this journey with you.