SevOne NMS Guidance on Availability & Backups

To create a SevOne NMS environment, a combination of HSAs (Hot Standby Appliances) with virtualization, snapshots, and an enterprise backup solution for the highest possible durability of network data is recommended. Additional options include SevOne Data Publisher or a custom backup solution created by Expert Labs or SevOne Support Team.

SevOne recommends all customers to consider virtual or cloud deployment for a number of reasons. Virtualization allows for higher availability and less down time for maintenance over the lifetime. It is trivial to move the Virtual Machines (VMs) to work around planned system maintenance and allow the move of SevOne VMs onto newer machines and age out the old machines. Doing the same with physical machines is possible, but takes considerably more time and effort, and more disruption to SevOne end users.

In addition to the capabilities that allow VMs to move between physical servers, it also simplifies backups. SevOne servers are backed up with a combination of vCenter snapshots and the Rubrik enterprise backup solution. Snapshots of all production critical machines, including SevOne NMS, are versioned and transferred offsite to the Disaster Recovery (DR) site. This also enables the restore capability.

SevOne offers an additional component to increase availability above and beyond the ability to use tools such as vMotion and VM snapshots. The HSA pairs with any peer in the cluster and receives all of its state in real time. Based on heartbeats or administrator action, the HSA can be promoted at any time from a secondary appliance to a primary appliance and take over the duties to collect data and respond to reporting queries. This reduces the risk if an appliance has an unplanned outage, or if a catastrophic disk failure were to occur that is beyond the capability of Redundant Array of Independent Disk's (RAID) setup to recover from. Customers who depend on SevOne typically take advantage of the HSA to expand their reliability and availability. SevOne Support Team has worked closely with the customers to recover from all kinds of failure scenarios utilizing the capabilities offered by the HSAs.

Some SevOne customers choose to continue to run SevOne software on bare metal, which has its own advantages and disadvantages. The ease of using vMotion and snapshots to push DR capabilities beyond what an HSA offers, are not possible with the bare metal. Most SevOne customers are satisfied with either a virtualized environment with access to their enterprise-wide snapshot and backup capabilities for DR purposes, or bare metal clusters with HSAs.

In the event that a customer is looking for even more durability of the data such as a cold backup, this presents a few challenges that are outside the capability of the SevOne product to solve.

SevOne NMS is designed to store one-year of data on disk, utilizing almost the entire space provided, which is considerable. Some SevOne customers across their cluster can hold over 1 petabyte (PB) of data. There is not enough space on the appliance to store a second copy of the data as an export / backup. In cases where customers work with Expert Labs to come up with a solution that works for their environment, it requires mounting external storage like ISCSI or NFS to be able to perform such exports. The exact implementation varies based on customer's environment and needs; and due to this, the solution cannot be productized.

SevOne’s core mission is to provide the best in class network performance functionality and continue to improve the capability, usability, and reliability of the software. SevOne's IT team uses the Rubrik enterprise backup solution however, there are several other solutions available. To solve this problem, it is outside the scope of SevOne product team’s charter, and needs to include incremental capabilities, be reliable in copying Terabyte or Petabyte of data across links that might not be sized for it, restart when connections are lost, consider the customer's business work hours so the WAN is not congested with backups while people are working, and assure that the backups retain integrity. It must be able to do all of this in reverse to enable a successful restore. Expert Labs has partnered with customers who have an existing enterprise grade backup solution to integrate SevOne with their workflow.

SevOne has reviewed the positioning of our two nearest competitors and found comparable guidance. The first vendor recommended snapshots / backups using the vCenter or cloud, and utilizing an enterprise-wide solution to manage moving those snapshots offsite. The second vendor offered a long list of manual configurations (not productized) to put periodic database dumps into a cron job. It did not resolve any of the issues stated above of how to ensure there is space for a successful backup, or manage its movement and lifecycle afterwards. It would certainly need to be customized to fit into the environment of each customer and their particular infrastructure and tools available.

Some SevOne customers achieve an additional tier of storing SevOne data by leveraging the SevOne Data Publisher. This capability can export all polled data to a Kafka broker, which then allows customers to connect to their Data Lake or Data Warehouse for additional analytics and long term storage. Customers assisted by Expert Labs have successfully performed the recovery by re-importing this data into their SevOne cluster.

SevOne has a number of capabilities that help customers increase the durability of their network data and allows customers to chose the best one based on their requirements. Some customers may find one of the above capabilities to meet their requirements, and others may feel that an engagement with Expert Labs to build a solution for their specific requirements to be the best choice.