Carbonite Availability software keeps critical business systems available and prevents data loss on Windows and Linux servers. Our continuous replication technology maintains an up-to-date copy of your operating environment without taxing the primary system or network bandwidth.
Quantitative Investment Management, Peak 10, KEB Hana Bank USA
We had applications that are file servers. The file server, and then we had the Oracle application. Then we have ERP software.
It’s the full server, physical server to virtual, physical to physical, virtual to virtual, like those things.
There have been benefits from using this product. For example, your server is down. You want to bring up the site in the shortest possible time. There is automatic failover, but I don’t prefer automatic because sometimes it may not be a hardware failure. It may be a network issue. That means if it goes for an automatic failover, then both servers will be running.
Therefore, what we prefer is manual failover on click confirmation, the remaining is automatic. You don’t have to do testing and all those things. Within an hour, you’re back online with the last committed data. Consider the server went down at 12:30, and your last saved data is 12 o’clock. That means your data is safe till 12 o’clock. But, again, you have the option to print.
So, within one hour of using this tool, it will be back online and running. Online means, consider, it is a failover that requires you to make sure everything is okay. You have to do your warehouse testing before giving it back to the production level. We are giving half an hour to one hour to make sure everything is running fine, like firewalls are running, those things.
Without that tool, without any surprises, it would take around a minimum of 24 hours to 48 hours if there are no surprises.
Like, you only went up the full server. You don’t know what packets have been running. Every time you’ll have some problems or a virus or something missing. This one is sort of already ready.
It was a very good product. It is real-time synchronization. It is a really beautiful product line.
Once it is synchronized, it is a bit-level synchronization across the VPN.
And the one beauty I really liked about the product line is you can run multiple jobs to multiple locations simultaneously.
Usually, whether it’s electronic or any other solution, you finish one job, and then it goes for the second replication job, not simultaneously.
In Carbonite Availability, you can have one DR service standby server in the head office, another one across the VPN at another DR site, and another one at a third site, like in the cloud. And all these three jobs will be going simultaneously once the initial synchronization is completed.
Like, if you delete here, it just deletes everywhere. That means it’s a bit-level synchronization. That beauty, I cannot find in any other product line.
It is real-time synchronization. It is high availability. That means if you delete something here, it deletes everywhere. That is the difference.
The only thing I didn’t like about this is when you do a failover or failback again. That was the only slow process. But otherwise, it is, like one way it runs beautifully.
I have been working from the start of the product line when it was Double-Take, and then we shared, and it was quite a long time I was working with that product line.
It is a scalable product. We have tested up to a terabyte of data, and it is doing very well. Initially, the first synchronization takes time. After that, it is almost immediate. Like, you delete here, you delete there. Even on a normal Internet connection, it is doing very well.
The scale up means, like, every year, how many physical servers you want to failover, but there are a lot of options. Like, one-to-one, one-to-many options. For example, you have five servers and you don’t want to invest in multiple servers. You want to keep one failover server, you can have that option, one-to-many. In that case, or many-to-one. Many-to-one means many servers will be the primary servers, and it goes to one other server. If any of the servers fail, then you can bring up that DNS server, that specific DNS server. That will be running on Hypervisor or VM.
I was in the enterprise segment, but for the last more than ten years, I’m in the SME segment.
Our local support is very good. Most of the time, we are able to sort out the issue. Once in a while, not much. We don’t require so much support because we take the pain of testing the solution in our lab and then make sure we understand the product line. So far, there is rarely a need to contact the back-end support, but they were really helpful also.
Positive
We have to size the servers properly. Like, when you do the staging servers and all those things, you have to do proper sizing of the hard disk and the memory. That is the only requirement.
If you are careful with that one, problem-wise, we didn’t face much of a problem. Once in a while, that’s the job. Because our consultant was not fully sizing the server.
So, whenever there’s a problem that doesn’t recover by itself, we are forced to delete the job and recreate the replication job. But, again, it doesn't take much time. That is the beauty of it. Restoring three or four jobs takes a maximum of ten to eleven minutes. That’s it.
Price-wise, it’s a bit expensive. But, again, when I compare it to the enterprise segment, it is not that costly because you understand what the solution you are taking is.
You’re not taking a backup solution; you’re looking at how much time you can wait for a return of service (ROS).
If a company is keen on return of service in the shortest possible time, then double-take. If they’re not bothered and can take a huge amount of time, like 24 to 48 hours, then this is not the product. But if you do a recovery like this from the DRS, it can have an offline backup also.
So, it depends on the customer's requirements. If the customer can wait some time, then that's good. If the customer cannot wait, then you must have something very strong. And this product gives exactly that.
The pricing is per server. If I’m right, it is around $4,000 per server license.
There are no additional costs to the standard fees. But, again, users can have it on-prem, and users can have cloud-based. The cloud will come with additional cost based on how much storage users are going to put onto the cloud and how much disaster recovery. That will add to the cost.
But on-prem, if users have done your hardware and how you utilize the hardware, storage is not very expensive. But the cloud storage prices are increasing. We can notice that. If you want good storage, good performance storage, you have to pay. But if it is on-prem, the prices will go down because storage prices will decrease.
Nowadays, it looks like this will be coming up more and more because we are moving into an electronic age, especially after the pandemic. That means customers cannot wait. Users need the application to be online every time, not like before. We cannot go through a manual procedure. This is certainly something everyone, even the SME segment, will look into.
It is coming as a hybrid solution, not a total cloud solution. Hybrid solutions are being picked up very well.
Ooverall, I would rate it an eight out of ten.
