• Documentation
  • Cloud-Edge Synchronization

Cloud-Edge Synchronization

Note

As the English version of the document is being updated, some of the content here might be inconsistent with the actual product or feature that you are using. Bear with us until we bring you the updated documentation in August, 2020.

As an extension of EnOS Cloud, EnOS Edge and Cloud can be synchronized under enough network bandwidth so that:

  • Automatic or manual synchronization of meta-data, such as models, assets, and applications, between Edge and Cloud
  • Real-time synchronization of device data to EnOS Cloud, such as device sampling data, statistics and alerts.

If the network bandwidth is not good enough, EnOS Edge offers the following functions:

  • Breakpoint resumption. This function caches data when the Edge is offline and use the following methods to transmit the cached data to EnOS Cloud when the network is resumed:
    • First-in-first-out mode: Data with older timestamps gets first transmitted.
    • Real-time-first mode: Data will be divided into real-time and historic data in their seperate channels. EnOS Edge makes sure that data in the real-time channel is prioritized and transmits the historic data only under condition that real-time data transmisstion is not interrupted.

Breakpoint Resumption

EnOS™ Edge supports breakpoint resumption. If the communication between the Edge and the cloud is interrupted, the Edge can continue to collect data from sub-devices and cache it locally. The data cached locally will be transmitted again once the network connection is restored.

The dual-way message confirmation mechanism for Edge and cloud ensures that the data cached in the Edge must be sent to the cloud, ensuring data integrity in case of network interruption.

The cache size can be calculated based on the following factors:

  • Number of measuring points
  • Data types
  • Sampling periods
  • Time-to-live period for cached data and
  • Storage size

The daily cache size per sub-device can be calculated as follows:

Daily data volume = (total measuring points × measuring point size (byte)) × (24h × 3600s/h ÷ sampling period (s))

Where the measuring point size can be defined as approximately 25 bytes.

Example

Assume that there are 100 devices, each sub-device has 50 points, each point and its related attributes occupy 25 bytes, the sampling period of the point is 5 seconds, and the data are continuously collected for 24 hours a day. Assume that a 64GB hard drive is selected;

Then the amount of data one day per device is: 50 × 25 × 24 × 3600 ÷ 5 = 21,600,000 bytes. So the total cached data for 100 devices in Megabyte (MB) is 21,600,000 × 100 ÷1024 ÷ 1024 ≈ 2060 MB

For the 64GB hard disk itself, space must be reserved for system, virtual memory space, and other required space. Assume that 40GB can be guaranteed for local storage;

It can store 40 × 1024 ÷ 2060 = 20 days of data.

Note

Excessively large data cache might cause issues in breakpoint resumption. Reserve a maximum of 4GB space for cache.