Build with an Azure free account. Get USD200 credit for 30 days and 12 months of free services.

Start free today

How to reduce inter-VM latency with Proximity Placement Groups

Play How to reduce inter-VM latency with Proximity Placement Groups

The Discussion

  • User profile image
    Jarmo

    Why you are running dsl commands on PowerShell console?

  • User profile image
    YanivShalom
    So basically you destroy the HA concept to get low Latency?
    meaning - if the Datacenter goes out you wont be able to reach your service...
  • User profile image
    ZivRaf
    @YanivShalom:
    Hello Yaniv
    Proximity Placement Groups are used for low latency either within a specific tier or across tiers in the app. They may not benefit all topologies, but should not impact you HA.
    Azure offers the following constructs for HA: Regions, Availability Zones (one or more DCs), and fault domains to spread instances (VMs).
    With fault domains (one or more physical racks) you are not guarenteed resiliency in the case of a datacenter level failure. Using PPG will not change that, but will force all the app tiers to the same DC for low latency. PPG is useful for multi-tiers app each spread 3 FDs. Without PPG, there tiers may be deployed across DCs in the region (and even in the AZ).
    In case you plan to span multiple AZs for HA, you may achieve low latency within each AZ by manually placing each VM in its AZ and PPG (having a PPG per AZ). This will work if you have an active-standby topology across AZs. If your app spread AZs in an active-active topology, PPG may not provide value for you.

Add Your 2 Cents