Find Communities by: Category | Product

VPLEX

9 Posts authored by: Parasar Kodati

Modern businesses are information businesses.  When technology fails, hospitals can’t do MRIs, factory production lines grind to a halt, and credit card transactions no longer go through.  Hence, the expectation for modern datacenters is that the technology cannot fail.  More than 50% of the Global Fortune 500 use VPLEX to deliver on this expectation because VPLEX has proven over the years to be synonymous with application availability.

 

Modern datacenters are also expected to keep up with fast changing business needs and to seamlessly integrate new technologies.  The rapid rise of flash as a storage medium is one example of the pace of change IT practitioners are asked to deal with.  Customers adopt flash for its simplicity and its performance.  At the end of the day, all-flash storage makes IT more productive and equally important, it makes IT’s business partners more productive.  That productivity translates into increased business competitiveness.  The business relies on that competitive edge to thrive.  But how does IT move their operations to all flash storage?  In many cases, time to value is measured in months of lost productivity gains and the migration process is complicated and expensive. Of course, more than 50% of the Global Fortune 500 companies who use VPLEX don’t have this problem because VPLEX streamlines the migration process and cuts down the time to value from months down to days/weeks.

 

EMC has just launched new lightning fast VPLEX hardware platform – the VS6 which is designed and built for speed, scale, and performance to meet the demands of tomorrow’s modern data center, embracing the trend towards deployment of all-flash storage arrays.


VPLEX For All-Flash with the lowest TCO

Recognizing the central importance of all-flash storage to the modern datacenter and to help customers accelerate their time to value, increase their agility, and deliver on non-stop operations, EMC is also offering a a new way customers can speed their adoption of flash and increase the availability and agility of their IT solutions – VPLEX for All Flash.

vplex4AF-new-rack-white.jpg

 

VPLEX for All Flash is an appliance model that combines VPLEX hardware and software at a single price per engine that includes an unlimited right to connect any number of EMC’s all-flash arrays for the purposes of protecting and moving workloads.  Any VNX Fx000, Unity AF, XtremIO, or VMAX AF capacity behind VPLEX is included.  When additional performance is needed, additional VPLEX engines can be added to the cluster. If there is a need to connect hybrid arrays, standard VPLEX licensing applies – frame-based for VNX/Unity, brick-based for XtremIO, and capacity-based for VMAX or third-party arrays.

 

VPLEX for All Flash summary

  • Pay per VPLEX engine, and place any EMC All Flash array(s) behind VPLEX
  • Available in Local and Metro versions
  • Available with VS2 or VS6 engines
  • Available in single, dual, or quad engine clusters
  • Paid Premium support is available on the appliance (HW and SW) at an attractive price point

 

If you are modernizing your datacenter with all-flash storage, consider VPLEX for All-Flash. It will accelerate your journey, simplify your availability, and bring new levels of agility that will make you wonder how you ever lived without it!

 

 

Author Bio

colin pic.png

Colin Durocher is a Consultant Product Manager with EMC and has spent over 10 years working on the VPLEX product in a variety of roles within Engineering and Product Management. He currently owns coordination with storage array partners, as well as pushing VPLEX performance and scale to meet increasing customer requirements. He’s also defining future VPLEX product strategy, direction, and innovation.

 

Outside of work, Colin enjoys racquet sports, brewing beer, camping, and spending time with his wife and two children in Montreal, Canada.

Last December we released GeoSynchrony 5.5 SP1 when we started supporting UNMAP for XtremIO for the basic Local case. Today we are glad to announce 5.5 SP2 with complete UNMAP support for XtremIO: Local Mirroring and Metro cases. The UNMAP support offers the following benefits to your VPLEX-XtremIO environment:

 

  • Reclamation of VM storage
  • Automatic reclamation of Windows Server storage due to data movement or deletion
  • Reduced the storage use on XtremIO backend arrays with VPLEX


This together with existing support for XCOPY, ATS, and WRITE_SAME, VPLEX now supports all of the VAAI primitives. Here is Itzik Reich, XtremIO CTO with more detail on the release. Also check out this great demo to see how UNMAP can be used to free up storage space that is claimed with thin provisioning through VPLEX.

 

We are very excited to announce VPLEX support for VMAX All Flash systems. Introducing an all flash array like VMAX All Flash is a great way of modernizing your datacenter. With data services matched by none and the all flash performance you are getting the best enterprise storage that can be the rock bed for your IT transformation efforts. Naturally, introducing VMAX All Flash into your datacenter means migrating hundreds of applications. Migration is a very important factor to consider if those workloads are using a storage array that is not VMAX. Storage tech refresh is one of the most daunting IT infrastructure projects that involves:

  • Having a migration plan involving pulling internal resources together
  • Planning the maintenance windows
  • Budget for professional services
  • Risk of things going wrong and losing data
  • Time to value for the new VMAX All Flash system

This is where VPLEX comes in. Enter online tech refresh. Nothing needs to go down, no more weekend maintenance windows and huge savings on that migration service bill.

gif_vmax_afa_tech_refresh.gif

Click on the image to see animation

 

Here is how VPLEX does it. VPLEX detects the new array, in this case VMAX All Flash, and copies the volumes from the old array to the new one until they are 100% in sync. Once in sync, VPLEX routes the IOs to the new array. At this point, after making sure everything is working properly, you can disconnect and retire the old array. While all this is happening, the host continues to read and write to the virtual volumes presented by VPLEX. With VPLEX in your environment you will never have to spend another weekend for storage tech refresh. VPLEX data mobility goes beyond the tech refresh use case to any mobility from patching servers to load balancing of storage arrays. With more than an Exabyte of data protected, moved and tiered, VPLEX is the most trusted solution when it comes to continuous availability and non-disruptive data mobility.

We previously announced the intent to move VPLEX Virtual Edition to End-of-Life (EOL). Today we are announcing the end of life effective February 8th 2016. All current VPLEX/VE customers will enjoy continued support for 3 years from the EOL date i.e., February 8th 2019.

 

EMC’s approach to managing product life cycle is based on serving our customers in the most effective way possible. It is with this commitment in mind that EMC has taking this step to end development on VPLEX Virtual Edition.  While VPLEX/VE is an innovative and unique product, EMC decided to better invest our resources in enhancing VPLEX core platform to better meet the demands of our growing market share. As part of the EOL efforts, VPLEX has removed the web content related to VPLEX Virtual Edition and the VPLEX Virtual Edition Trial.  Customers will continue to have access to VPLEX/VE related documentation and downloads at support.emc.com.  The VPLEX family would like to thank all of you for your interest, and we look forward to improvements to the VPLEX core platform in the future.

We are very excited to share with you VPLEX CloudArray integration that extends the key VPLEX use case of mobility to public and private cloud storage. The use case for cloud storage behind VPLEX hinges on frequency of use: cold data that is accessed much less frequently can be pushed to volumes on CloudArray and its extension to public and private cloud. This saves valuable primary storage and helps you take advantage of more cost effective cloud storage for cold data. More details in this great white paper by Don Kirouac.


vplex cloudarray.png

Following are some use cases that show the need for varying frequency of data access:

 

Splunk host/warm/cold/frozen tiering - CA to provide frozen/cold tiers

Splunk is an application that collects logs from datacenter equipment and enables query functionality on those logs.  It collects enormous amounts of data over time.  Since the more recent information is generally more interesting to queries / analytics, Splunk has the ability to tier data (typically by time) into hot/warm/cold/frozen tiers.  The frozen tier is not included in any queries at all unless it is promoted back to cold. It is possible to store cold / frozen tiers on CloudArray so that faster and more expensive primary storage arrays are freed up.  They would only have to serve the hot/warm tiers in this case.

 

Microsoft Exchange archive mailboxes

Due to cost concerns, users are often provided mailboxes which are too small.  Users can then use archive mailboxes to store older emails for posterity and this is common practice.  Exchange even allows one to setup policies to automatically move older emails to an archive mailbox. With CloudArray, it is possible to host the inbox on primary storage and archive mailboxes on cloud-based storage for lower cost.  Since archive mailboxes are accessed less frequently, CloudArray provides adequate performance to host these.

 

Oracle and DB2 Database partitioning by key (i.e. timestamp)

Databases such as Oracle 12c and DB2 offer the capability to partition tables by a primary key – which could be a date.  In this way, older data can be stored in a separate table which can live on slower storage – such as CloudArray.  Since it is touched less frequently by database queries and since the database queries that do access these records tend to be Decision Support workloads (sequential in nature) the performance of CloudArray is appropriate for hosting this older data.

 

SQL Server snapshot / backup long term retention, manual tiering

Database administrators often use snapshots for operational recovery of their databases.  The older the snapshot, the less likely it is to be used.  With VPLEX and CloudArray, it is possible to migrate snapshots from the primary array to the Cloud so that it can be deleted from the primary array.  This frees up space on the primary array and lowers the total cost of keeping these various copies of the database.

 

Feel free to share how you plan to take advantage of the cloud tiering capability of VPLEX. We are eager to know.

UPDATE: Complete UNMAP Support in SP2 available from 4/22/2016: Complete UNMAP support for XtremIO

 

Thin provisioning is a great way of getting the most of your valuable storage. This is especially important in case of XtremIO which is an all flash array of highest performance level. With 5.5.1 VPLEX supports VMware VAAI UNMAP command. The newly added support for UNMAP allows VPLEX to free unused blocks from thin volumes provided by XtremIO arrays.  With this capability, you can now reduce the size of thin volumes when you move VMs or delete content from ESXI hosts and even within certain VMs.  The ability to use UNMAP is added to the already existing support for VMware VAAI commands like ATS, XCOPY, and WRITESAME.

 

unmap xtremio local.png

For the VPLEX 5.5 SP1 release, the use of UNMAP is limited to XtremIO arrays and VPLEX Local volumes only. At this time VPLEX Metro volumes, and VPLEX volumes with local mirrors are not yet supported. Additional support for UNMAP on XtremIO array and VPLEX will be released in follow on versions of VPLEX software. Stay tuned!

With GeoSynchrony 5.5 Patch 1, we are really excited to announce a major scale up of the MetroPoint topology to 4500 virtual volumes (more than 200% increase!). This enhancement in scale is very much in line with VPLEX product strategy of being focused on the core mission of eliminating planned and unplanned downtime in your environment.

 

Click on the image for animation

metropoint animation.gif

Click on the image for animation


We launched the MetroPoint configuration in 2014 as a unique three site deployment of VPLEX Metro together with RecoverPoint. MetroPoint is more than sum of its parts: VPLEX Metro offers continuous availability over metro distances and RecoverPoint offers continuous remote replication over long distances for disaster recovery. In a MetroPoint topology a single DR site/copy protects both the metro sites. The same DR copy can be used to recover either of the metro sites thereby providing resiliency to entire regional disruptions. Many of VPLEX's largest enterprise customers from Telkom Indonesia to Societe Generale to Dow Corning have protected their continuous availability environment with the long distance DR site protection in a MetroPoint topology.

Monitoring performance is very valuable to IT infrastructure staff to troubleshoot potential performance issues or take actions like balancing workloads across the storage arrays that are behind VPLEX. We are really excited to announce the first release, v1.0 (release notes), of the new VPLEX Performance Monitor, a performance monitoring tool that gives users visibility into virtual volume performance (no matter which array it is on) for troubleshooting and analyzing trends in IO traffic going through VPLEX. The tool is free with VPLEX GeoSynchrony and can be downloaded here. Check out this great demo:

 

 

VPLEX Performance Monitor is a new stand-alone feature designed to give users the visibility into how their environment is performing for the purposes of debugging and trend analysis. This will allow customers to review the performance of their volumes over long periods of time, up to 30 days, and to inspect specific performance at targeted times. The tool is customer installable and a configurable VM.

 

VPLEX Performance Monitor charts the following virtual volume performance metrics and allows users to zoom in and out of regions of interest:

  • IOPS
  • Read Bandwidth
  • Write Bandwidth
  • Read Latency
  • Write Latency

VPLEX hits a key milestone

If there is one solution for application continuity as elegant as an uninterrupted power supply is for power continuity, it must be EMC’s VPLEX. My colleague Colin Durocher recently shared our laser focus on Continuous Availability when it comes to product strategy and here are some stories our customers shared with us when VPLEX saved the day, thanks to that focus. Today I am very excited to announce that VPLEX hit the 100 million hours of operations, more than 10,000 years of keeping mission critical applications always on!

 

We are also announcing GeoSynchrony 5.5 (software that powers VPLEX) with some great feature additions for ease of use, performance and scale enhancements and VMware SRM integration:

 

Easy provisioning for VMAX3 and XtremIO

VPLEX Integrated Array Services (VIAS) functionality cuts down storage provisioning from a laborious 45 step process across multiple UIs to 5 simples steps within VPLEX Unisphere UI. With GeoSynchrony 5.5 this simplified provisioning of VIAS is extended to:

VMware SRM integration

VPLEX works seamlessly with VMware stretched clusters to deliver cross datacenter business continuity and mobility. With 5.5 the active-active deployment is extended to VMware's Site Recovery Manager (SRM) through a VPLEX SRA. This integration means ability to meet stringent SLAs for business critical applications and data:

  • Zero RTO/RPO for planned failovers
  • Zero RPO for unplanned failovers

For additional protection against data corruption they can additionally deploy RecoverPoint resulting in a MetroPoint topology.

Performance and Scale

With GeoSynchrony 5.5, VPLEX delivers improved performance for transaction intensive workloads with Optimal Path Management—optimally assigning virtual volumes to directors thereby reducing latency and increasing IOPS. Performance improvements:

  • Up to 25% faster for Local
  • Up to 10% faster for Metro

Note: This performance enhancement is targeted primarily for transaction intensive workloads and is available upon request.

 

There are also scale improvements that matter especially with a highly scalable enterprise platform like VMAX3. VPLEX now supports up to 10,000 Local Virtual Volumes. Our scale support for 8000 Metro DR1s remains at 8000 but with improved usability in both the UI and CLI. Overall user experience has been improved for the UI and the CLI for day to day maintenance and management of clusters (Local and Metro). Please refer to release notes for all supported cluster environments.

 

Coming soon! Stay tuned to learn more about all new 30-day Performance Monitoring tool! Please note that the new tool will be available as a separate build package installable independently on a VM and not on part of GeoSynchrony 5.5.

 

Last but not the least

If you have not checked these already our technical team has put together some great best practices videos covering a wide range of topics from provisioning to load balancing:

 

Basic VPLEX storage provisioning

VPLEX VPN Troubleshooting

Using solve desktop for VPLEX

FrontEnd load balancing (Choosing the right multipath software)

VPLEX Virtual Volumes

Resolving VPLEX time drift issues

Filter Blog

By date:
By tag: