Start a Conversation

This post is more than 5 years old

Solved!

Go to Solution

525

October 15th, 2014 08:00

Cloud & Object Storage Solutions

How are ViPR Services different from other unstructured, cloud and object storage solutions in the market?

64 Posts

October 15th, 2014 12:00

The real differentiation of the ViPR Services is its scalability, performance, and software-only design. Some key differentiation include:

• Data is stored in append-only containers: ViPR stores data in “chunks” (aka containers). Containers are append-only, once a container is full, it becomes immutable. This also enables support for multiple concurrent writes to an object. First, a write updates the container, and then updates the index. Only after the index is successfully updated will the write be successful. Over time, authorized deletes may result in a container which has very few indexed objects in it. There is a garbage collection process that will identify a container with a large percentage of objects that are no longer in the index. In this case, ViPR can create a new append-only container, move the objects and update the index. This ensures very efficient storage.

• Geo-scale index: A high-performance, distributed indexing mechanism to find data very quickly.

• Geo-scale metadata: ViPR stores metadata associated with customer data separately from data. It’s stored in a very efficient way so we can efficiently run operations against the metadata.

• Low-latency transactions: ViPR can execute a large number of user transactions concurrently with very little latency. ViPR supports box-carting to handle workloads with high transaction rates. When an application is writing a lot of small files with high I/O, ViPR can take multiple requests together and write them as one. This improves performance by reducing the round trips to the underlying storage. Competitors market similar capabilities but require a separate platform such as SSD; whereas ViPR executes in software. Box-carting is only enabled if the request is smaller than 512 KB, and there are high numbers of requests per second.

In addition to the data, the index and metadata are located on the actual arrays in containers. Consequently, the index can scale massively and be geo-distributed. By storing the index and metadata on the actual arrays, ViPR Object can locate objects in a massive scale, geo-distributed object store much faster.

No Events found!

Top