If you are not interested in using an online backup solution than this post isn’t for you. However, if you are evaluating or looking to change online backup solutions and specifically trying to evalutate the Ongoing Operations’ Replicator solution vs. Comvault you are at the right place. The goal of this post is to provide general education of each criteria and than a chart at the bottom comparing the Replicator to Comvault.
What will it backup?
One of the first criteria is determining what the solution will backup. Specifally what operating systems and types of data (database vs. email vs. application) will you need to backup. Key questions to ask are:
- What operating systems are suported? Windows, Linux, Unix
- Does the software care what is running on a server (database, exchange, application) or does it just backup the whole thing?
- Does the solution support Physical or Virtual machine backup and recovery?
Performance of the Backup
This is an essential question that will provide you with one of the most essential pieces of information. There are essentially two types of backups that can be perfromed – File and Block. File level backup is where you take whole copies of specific files or directories at some time interval. Block level is where you are recording the changes at the disk (hard drive) level at some time level. Your other options are Full backup or Incremental. Full backup takes an entire copy of all the data each time and incremental takes the changes from the last time a backup was performed.
Local Restoration vs Offsite
Another key question is determining if the solution can key multiple copies of the data. You will probably want a copy locally for small – single server recoveries and you will want a copy offsite for major recoveries. Some key questions are:
- Does the solution provide local and offsite copies?
- Where does the data go for the offsite version?
- What is the security level at the offsite location?
- Does the solution provide local hardware for recoveries?
Once you have local copies of your servers or data you will than be concerned with how you get the data offsite. Specifically:
- Can I throttle and set the amount of available bandwidth to allocate for data transport?
- Is the data encrypted during transport?
- How will I know if the data didn’t get offsite?
Data archival is an important question. You want to know how long your backups are going to be kept and what your options are for recovering something from several years ago. The key questions are:
- How are historical snapshots of your data kept?
- How long are they maintained?
- How much storage will you need to maintain the backups long term?
- Can the solution perform Data Deduplication? Essentially meaning avoiding storing the same file that never changes 365 times.
Protection of the data at rest
Encryption of the data is all about making sure that the data is protected from either someone getting into your solution or from your data leaving your environment. The key is to make sure you know:
- What tools does the solution have to make sure nobody can access the data other than you?
- Is the data encrypted first and then transported or is a backup created and then transmitted?
- Is it encrypted locally and offsite?
- Who sets the encryption key?
Management of the solution
Management of the solution should be a key criteria and you should pick a solution based on how you actually work normally. Ie. Do you like to learn the application and know it inside and out or are you comformatable outsourcing some knowledge that you probably won’t need very often. Key questiosn are:
- How much time and effort do you, the client, have to put into the solution?
- Does the vendor constantly monitor your backup operation to ensure performance?
- How do you know if the data got offsite or if a backup worked the night before?
- Do your staff need to be trained on the solution to run it or is mostly co-sourced?
In the past, most backup solutions required you to rebuild the machine (reload windows for example) and then recover your data from the backup solution. This was extremely inefficient. Most solutions today should allow you to perform Bare Metal Resotores, or restoring the data and the Operating System directly from the backup at one time. This can save lots of time. So the key comparison here is:
- Support for Bare Metal Restores or File level Recovery?
- Recovery from a full backup and then restore incrementals or Point in Time recovery?
The need to test the solution can be a huge component of the time you spend on the product. Generally our clients have to test at least once a year and some 4 times a year. Making sure the platform makes it easy to recover a server is crucial. It is also important to deterine if you or the vendor will be conducting the test as this factors greatly in to the TCO, total cost of ownership.
- Who will provide the test environment and what happens if you don’t have enough space?
- Who will conduct and certify the test?
- Who will troubleshoot problems that occur?
What happens in an actual recovery?
- Are you paying the vendor to simply take snapshots and move your data or are they helping you actually recover the servers?
- Do they provide the recovery platform?
- Do they connect you to critical third parties?
Want to learn more? Fill out the short form below and a member of the OGO team will be in touch shortly!