It was the 20th of July when I saw the following message on southwest.com while looking to book a ticket for an upcoming customer visit. My immediate thought was that a critical piece of the infrastructure had failed and their backup and recovery processes are going to be exercised in full gear. But when the site message persisted for a few hours, I realized that it is probably going to cost Southwest millions of dollars in lost business, penalties and compensation. Eventually it was announced that the downtime was due to a bad router and the failure of a redundant backup system. With more than 2000+ flights either delayed or canceled, it took 12 hours to reboot 400 computer servers to get the system operating. That’s an RTO of 12 hours in this case and for businesses that are looking to reduce RTO SLAs of their backup system, Cohesity can help with Cohesity DataProtect which is powered by our patented SnapTree technology.
Before we get into how Cohesity SnapTree significantly lowers the RTO, let’s look at data protection in legacy storage solutions. These solutions leverage copy-on-write/redirect-on-write snapshot technology to create copies of data. Over time, these snapshots form a chain by tracking the changes made to a set of data. Every time a change is captured, a new link is added to the chain. As these chains grow with each and every snapshot, the time it takes to retrieve data on a given request grows because the system must re-link the chain to access that data. This process is highly inefficient and leads to high RTO windows.
Cohesity SnapTree technology is a ‘Distributed-Redirect-on-Write’ (DROW) snapshot mechanism that provides speed and scalability in backup and recovery processes. All nodes participate in this process thereby leveraging the distributed scalable elements of the Cohesity architecture.
Cohesity’s DROW implementation keeps track of changes by writing the changed data to new blocks. SnapTree then creates a tree of pointers that limits the number of traverses it takes to retrieve blocks of data, regardless of the number of snapshots that have been taken. With SnapTree, access to any point in the tree is fixed no matter how many snapshots there are and does not involve rebuilding any chain linkage. Furthermore, the snapshots are fully hydrated leading to fast recovery times. Get ready to board – Cohesity for Data Protection and low RTO @ gate SnapTree.
