Data Archiving


Data archiving, also known as cold data dumping, refers to the migration and archiving (dumping) of some very old and low-access data to inexpensive storage systems such as Amazon S3.

Starting from version 1.14, SensData provides data archiving function to reduce the cost for customers.

When do you need data archiving

Data archiving can be performed under the following conditions:

  1. When users have been using SensData for a long time and the data stored in the cluster is increasing, requiring regular expansion of the cluster's storage size
  2. When users have been using SensData for a long time, and the access frequency for older data is low

Expected effects of data archiving

After data archiving, the following expected effects will be achieved:

  1. A large amount of old data will be stored in cheaper storage systems, reducing the cost of use.
  2. Archived data will not affect the query of old data, but the query speed will be slightly slower than storing it in the original cluster.
  3. The date that has been archived can no longer be imported.

Preliminary understanding

Currently, data archiving provides three different storage protocols:

ProtocolDescriptionUse cases
hdfs://Hadoop cluster file systemIf there is a low-cost HDFS storage environment, this protocol can be used for data archiving.
s3a://Amazon S3 cloud storageIf you establish an archive file server by directly creating an Amazon S3 bucket
jfs://JuiceFS SDK access to cloud storage (pay-as-you-go)Access various supported cloud storage through JuiceFS SDK


Reference documentation:

Enable function

To enable data archiving, please contact the support team of Sensors Data. The Sensors Data technical team will provide the most reasonable archiving solution and enable archiving service.

How to delete archived data

If you need to delete archived data, you cannot delete it directly on the NAS disk. You need to use the Data Cleaning Tool.

The delete command cannot specify an event name, otherwise deleting the data will result in an error. Once the archived data is deleted, the deleted data cannot be queried again, and the start date corresponding to the start of the event in Sensors Analytics will automatically adjust.