EMC World – Session Notes – RecoverPoint: Accelerated Recovery in a Virtual Environment

Fantastic session….very good content and great demos.

4.0 Features

  • Virtual RPA’s
  • 7x capacity
  • Fan-in 4:1 or 1:4
  • SRM any point in time support with RP
  • New Unisphere looking RP GUI
  • Sync replication over IP
4.0 Themes covering in this session – Virtual RPA, Simplified UI, Any PiT with SRM, REST API, Hyper-V CSV Support, Multi-Site (4:1, 1:4) w/8 kb protected LUNs
 
4.0 Themes not covered in this session – eLicensing Self-Service, 2 PB per cluster, Sync over IP, Customer Upgradeable, Application QoS, 400 MB/s async per application, FIPS 140-2
 
More after the jump…

Why RP for VMware? Local & Remote, Any Storage, Scales to 1000s of VMs, Sync & Async, Physical & Virtual, VAAI

Can do cross-array consistency groups – very cool – SRM is aware of this.

SRM Point in Time Capabilities
  • Components
    • vCenter 5.0 and up, SRM 5.0 and up
    • RP & SRA
    • Config: new CG policy
    • Operation: VSI 5.5 (base and plugin) to select points in time – fully managed by vCenter
  • 3 main uses cases
    • 1. DR Test with Any PiT
      • Why? Test earlier PiT, App-consistent PiT, re-test same PiT, re-test with different PiT
      • Re-testing same PiT is really good to me.
    • 2. Operational Recovery to any PIT
      • Retrieve data from 30 minutes ago
      • Can do this with RP but SRM makes it much easier to use this method for day to day recoveries
      • Use SRM to bring app up and then extract data however desired.
      • THIS IS VERY COOL.
    • 3. Actual Failover to DR site
      • Will my failed-over image work?
      • Straightforward
  • Demo of UI
    • Unisphere UI for RP is quite nice.
    • Make bookmark for RP, then use VSI in vCenter to “apply” whichever PiT you want, then it’s just regular SRM interface.
Virtual RPA
  • No cost – just pay for software.
  • Need OE 5.32.000.5.201 and an iSCSI SLIC
  • Need to have iSCSI connectivity between VNX & ESX
  • Each vRPA should have 4-8 GB RAM, 2-8 vCPUs, 3-20 GB Ethernet
  • When physical vs. virtual?
    • Virtual for low-end workloads for low-end VNX workloads
    • One vRPA tested to sustain 8500 IOPs, 80 MB/s throughput
      • Number based on 8 vCPUs, 8 GB RAM (no details on processor type)
    • Performance is linear in the number of vRPA’s
    • Use EMC’s planning tools
  • Demo of vRPA – takes about 40 minutes including progress bars (demo was 5 minutes). Very easy…
Hyper-V Support
  • Failover with Hyper-V & CSV
  • Support for ODX – Offload Data Transfer
  • ODX supported with 05.32.000.5.21 for VNX splitter
  • ODX supported with any RP release that supports R32
  • ODX supported between two arrays actually – interesting.
Roadmap Stuff – no commitments
  • If have virtual appliance, why not embed it into the array?
    • 3 options – physical RPA, vRPA in ESX, vRPA in the array.
Questions
  • FC capable
    • Yes, can split FC workloads to the vRPA via iSCSI but just have to keep ghroughput in mind – i.e. 10 Gig iSCSI
  • any offload in physical RPA
  • processor type for the vRPA spec?
    • 12 core Sandy Bridge

One thought on “EMC World – Session Notes – RecoverPoint: Accelerated Recovery in a Virtual Environment

  1. Pingback: Think Meta » EMC World 2013 – Blog Posts Reference

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s