Skip to main content
  • Place orders quickly and easily
  • View orders and track your shipping status
  • Enjoy members-only rewards and discounts
  • Create and access a list of your products
  • Manage your Dell EMC sites, products, and product-level contacts using Company Administration.

Dell PowerFlex 4.6.x Technical Overview

PDF

SAN virtualization layer

This section details the steps for exposing the virtual SAN devices to servers with installed and configured SDCs.

The MDM cluster manages the entire system. It aggregates the entire storage that is exposed to it by all the SDSs to generate a virtual layer - virtual SAN storage. Volumes can now be defined over the storage pools and can be exposed to the applications as a local storage device using the SDCs.

To expose the virtual SAN devices to your servers (the ones on which you installed and configured SDCs), perform the following:

  1. Define volumes. Each volume that is defined over a storage pool is evenly distributed over all members using a RAID protection scheme. By having all SDS members of the storage pool participate, PowerFlex ensures:
    • Highest and most stable and consistent performance possible
    • Rapid recovery and redistribution of data
    • Massive IOPS and throughput

    You can define volumes as follows:

    • Thick

      Capacity is allocated immediately, even if not actually used. This can cause capacity to be allocated, but never used, leading to wasted capacity.

      Thick capacity provisioning is limited to available capacity.

    • Thin

      Capacity is “on reserve,” but not allocated until actually used. This policy enables more flexibility in provisioning.

      Whereas thick capacity is limited to available capacity, thin capacity provisioning can be oversubscribed, as follows:

      Maximum thin capacity provisioning = 5 * (total raw capacity - used capacity)

      When capacity usage reaches the level where it may cause I/O errors, alerts are generated. At certain higher capacity levels, volumes (even thin volumes) can no longer be created.

      Example:

      In a system with 3 SDSs, each with 10 TB, there are 30 TB of storage.

      In the system, there is already a thick-provisioned volume that takes up 15 TB of the total raw capacity (created by adding a 7.5 TB volume).

      MDM will allow a total of 300 TB total raw capacity to be provisioned, and since 15 TB are already allocated, you can add a thin-provisioned volume of 285 TB total raw capacity (by adding a 142.5 TB volume) or a thick-provisioned volume of 15 TB total raw capacity.

      NOTE:This example uses 10x fine granularity over provisioning plus compression support.
  2. Map volumes. Designate which SDCs can access the given volumes. This gives rise to the following:
    • Access control per volume exposed
    • Shared nothing or shared everything volumes

      Once an SDC is mapped to a volume, it immediately gets access to the volume and exposes it locally to the applications as a standard block device. These block devices appear as /dev/sciniX where X is a letter, starting from “a.”

      For example:

      • /dev/scinia
      • /dev/scinib
    • The maximum amount of partitions for the scini disk is 15.
  3. The maximum amount of volumes that can be mapped to an SDC is listed in the “Supported capabilities and system limits” table.

    NOTE:SDC mapping is similar to LUN mapping, in the sense that it only allows volume access to clients that were explicitly mapped to the volume.

Rate this content

Accurate
Useful
Easy to understand
Was this article helpful?
0/3000 characters
  Please provide ratings (1-5 stars).
  Please provide ratings (1-5 stars).
  Please provide ratings (1-5 stars).
  Please select whether the article was helpful or not.
  Comments cannot contain these special characters: <>()\