The new features include:

API-Only:

• Container Quotas
• TempURL File Name Overrides
• Bulk Operations (Delete & Auto-Expand Archive)
• Static Large Object support

Now in Control Panel:

• Multiple File Upload
• Editing Headers

All of these features are now available in both the U.S. and U.K. regions.

To avoid the risk of getting into TL;DR territory; let’s take a high-level look at how each new feature works. I’ll follow-up with subsequent posts in coming weeks that dig deeper into how each new feature works along with specific API details.

Here’s what’s new:

Container Quotas

You can now set simple quotas on your containers by setting the X-Container-Meta-Quota-Bytes and X-Container-Meta-Quota-Count headers in your container metadata. This can be useful for administrators who want to put limits on containers used by non-admin users (users who cannot modify container metadata). The X-Container-Meta-Quota-Bytes header is used to limit the maximum size of a container, in bytes. The X-Container-Meta-Quota-Count header is used to set the maximum object count of the container.

TempURL File Name Overrides

With the new TempURL File Name Override feature, you can specify a “filename” query parameter to override a TempURL’s Content-Disposition header in order to instruct their browser to save a TempURL file with a specific name. This is done by appending &filename=my_file_name to the end of a TempURL.

Bulk Operations (Delete and Archive Auto Extraction)

Two new bulk operations are now available: delete and archive auto extractions. You can upload and auto-extract tar archives upon PUT request to your Cloud Files account. This is done using the query parameter ?extract-archive=format specifying the format of archive file. Accepted formats are tar, tar.gz, and tar.bz2. Root directories in the archive will be converted to containers if uploaded at the account level, or mapped to pseudo-directory names if uploaded into existing containers.

Static Large Object Support (SLO)

Static Large Object Support (SLO) is very similar to Dynamic Large Object (DLO) support in that it allows you to upload many objects concurrently and download them afterwards as a single object. It is different than DLO in that it does not rely on eventually consistent container listings to do so. Instead, a user-defined manifest of the object segments is used to help guarantee that every segment in the manifest exists and has matched the specifications. If everything does match, you will receive a 2xx response and the SLO object is ready for downloading. This can improve consistency when uploading multiple segments concurrently and subsequently downloading the resulting large object a short time after.

Multiple File Upload

You can now use the Control Panel to select and upload multiple files through one Upload Files operation.

From your container, select the Upload Files operation and the click the Choose Files button.

Select the multiple files you want to upload to your container.

After you have selected them for upload, notice that the number of files selected is shown next to the Choose Files button. Now select the Upload File button.

At this point your files will be uploaded into your Cloud Files account. Once the operation has completed, the Control Panel will show the updated contents of your container.

Editing Object Headers

You can now also use the Control Panel to modify some of the most common HTTP object headers including:

• Access-Control-Allow-Credentials
• Access-Control-Allow-Methods
• Access-Control-Allow-Origin
• Access-Control-Expose-Headers
• Access-Control-Max-Age
• Access-Control-Request-Headers
• Access-Control-Request-Method
• Origin
• Content-Encoding
• Content-Disposition

To edit object headers, go to the cogwheel next to your object and select Edit Headers from the drop down.

Select the header you wish to set along with its value and click the plus sign to add:

You will see any custom headers that have been set when you select the View File Info dropdown from your object’s cogwheel. Reselecting Edit Headers from the cogwheel drop down will allow you to set any additional headers or modify existing ones that you had previously configured.

For additional information on setting object headers or details on any other Cloud Files features, please refer to the Cloud Files API Developer guide.

Stay tuned for upcoming posts with details on how to access these new features through the API with command-line examples using cURL.

If you have additional questions, please feel free to ask them here, contact our Fanatical Support Team or email me directly at jose.malacara@rackspace.com.

Be sure to subscribe to our product and development blog for future product updates and announcements.

For a deeper look at some of these features, head over to the Rackspace DevOps Blog.

Two Flavors To Pick From

Rackspace Cloud Block Storage has been designed to serve two very distinct use cases:

1. customers who need to scale their attachable storage independently from their Cloud Servers
2. customers who need high performance storage for databases or I/O intensive applications

Because these use cases have different needs, Rackspace built two solutions that make up the Cloud Block Storage product. The basic performance tier gives customers that simply need more attachable storage a low cost solution based on commodity drives. The high performance tier will allow customers to take advantage of speedy solid-state drives.

How Rackspace Cloud Block Storage Works

The diagram below illustrates the Cloud Block Storage architecture:

For Cloud Block Storage, we started with the Nova Volume API of OpenStack as the core interface. We then created an internal storage system called Lunr as the storage system behind the Volume Manager API. This storage system builds on top of standard Linux technologies and commodity hardware. Lunr integrates with OpenStack Nova through a nova-volume driver. Cloud Block Storage API requests are made through the OpenStack Nova Volume API and passed on to the Lunr API.

Volumes are provisioned in Lunr and attached to the Cloud Servers Compute host via iSCSI. The host then exposes the volume to the guest as a virtual device. Backups of the volumes can be made as snapshots and are stored in Rackspace Cloud Files (Swift Storage). Our storage nodes can be populated with either high performance solid state drives (SSD) or standard hard drives.

Features & API Operations

Cloud Block Storage uses the following API operations to attach volumes to Cloud Servers powered by OpenStack and manage the volume attachments:

Simple Pricing

We built a unique and simple pricing approach for Cloud Block Storage. Unlike other cloud providers, our simple pricing model is not based on input/output operations per second (IOPS), since the vast majority of businesses don’t know and often can’t control their IOPS. We will offer a flat rate per GB of storage. Final per GB prices are not set yet, but will be very competitive.

Currently In Preview (Beta) Testing

Rackspace Cloud Block Storage is now in Beta/Preview stage which means our Engineering and Product teams are gathering feedback, tweaking system performance, fixing any bugs and perfecting our Fanatical Support for the product.

We’ve had an overwhelming response to the CBS Preview so we are asking customers (which have been invited to the Preview) to complete a survey and we are using the survey responses to prioritize a good mix of use cases, performance testing, etc. We still have a number of customers in our queue to get access to Cloud Block Storage Preview and may invite more customers if needed.

We are very excited about this offering and we’re working hard to launch it into Unlimited/General Availability soon after the August launch of our open cloud based on OpenStack.