We use it for disaster recovery and to migrate machines from one location to another.
We use it for disaster recovery and to migrate machines from one location to another.
The big thing for us was our disaster recovery. At that point, we were only able to do a disaster recovery test once a year. Now, we officially do a disaster recovery test once a quarter and we do a subsequent test once a month to verify that it's doing what it's doing and the IP address is changed. Instead of one mass disaster recovery exercise, we're easily able to perform up to 12 in the year.
It allows us to verify in a much more granular aspect whether our data is being migrated or not. Once a year, if we find some issues, we're at least 11, 12 months behind at that point. Every 30 days, if we do a test and we find an issue, we're able to correct that. The time between tests is shorter, which means that if there is a problem we're able to resolve it in a much shorter amount of time versus an entire year, and then waiting another year to see if everything is working again.
When we need to failback or move workloads Zerto decreases the time it takes and the number of people involved. We are able to put a machine into Zerto, let it do its magic in migrating the data from one side to the other. We've had instances where we've got machines that are four or five terabytes that we can move from one side to another after it's done synchronizing in 15 minutes or less. Sometimes it takes DNS longer to update than it does for us to move the machine.
Instead of me having a server person, a network person, and a storage person, I can put it into Zerto, let Zerto do its job, fail it over, and then just have the application owner verify that the server is up and running, and away we go. So on a weekend, I don't have to engage a team of people, it can just be myself and one other person to verify that the machine is up and running. It really cuts down on overhead for personnel.
In situations of failback or moving workloads, it saves us hours. If I were to have to move a four or five terabyte machine using something like VMware's virtual copy it has to install on the machine and copy the data over. Then it has to shut the machine down and do a final copy, which means there's a lot of downtime while it's doing the final copy. As far as downtime from an application standpoint, with Zerto, we're from hours down to minutes, which is great when you have applications that are supposed to be the five nines of a time kind of thing.
We have not had any ransomware issues. But we have had an instance where somebody installed something that messed something up. It was a new version of Java and we were able to roll back. Thankfully they realized it fairly quickly because we only keep a 12-hour window. We were able to roll back to almost a per minute instance prior to that installation and recover the server in minutes. Our backup was as of midnight, but they did it at 8:00 in the morning. So we didn't lose eight hours' worth of processing.
If we were going to use our backup solution, it would have taken minutes to restore the actual server, but then from an SQL perspective, we would have had to roll the transaction logs from backups. I couldn't even tell how long that would have taken because we had to do all of the transaction logs, which are taken every five minutes from midnight, all the way to 8:00 AM in five-minute increments. It would have taken considerably longer using traditional methods versus Zerto.
Although it hasn't reduced the number of staff involved in overall backup and disaster recovery, what it has allowed us to do is actually focus on other things. Since Zerto is doing what it's doing, we're able to not have to stare at it all day every day and make sure that it's working. We have the screen up to make sure there are no errors, but we're able to focus on learning how the APIs work, working on the other products that we own for backup and storage. That's mainly what my group does, we do disaster recovery and storage backups. We have six pieces of our enterprise and before it was just the main piece that we were working on. Now, we're able to actually work with the other five or six entities and start doing their backups and disaster recovery because we have a lot more time.
The failover capabilities are definitely the high spot for us. Previously, when we did disaster recovery it would take us easily a day or two to restore all of our servers. We can do the same thing with Zerto in about an hour and a half.
We're about six or seven seconds behind our production site and it does a really good job of keeping up, making sure that we're up to date. That's one of the other things that we think is just phenomenal about the product, we're able to get in there and put a server in and within usually a few minutes we're protected. Six or seven seconds behind is a pretty good RPO.
Currently, we are using another product for longterm retention, so I don't think we really have any plans on switching over at this point.
Zerto is very easy to use. We did a proof of concept and it took longer to build the Windows servers that had to be installed than it did to actually install it and roll off the product. Our proof of concept became production in minutes.
The interface is the only thing that we've ever really had an issue with. It's gone through some revisions. The UI, it's not clunky, but it's not as streamlined as it could be. Some of the workflow things are not as nice as they could be.
I like the fact that Zerto does what it does and it does it very well. I have had Zerto since version four, so the longterm retention and things like that were never a part of it at that point. I just like the fact that I can install it, I can protect my virtual machines, and I'm comfortable and confident that it's doing things correctly because of the amount of testing that we've done with it.
We have been using Zerto for a little over three years.
It's very stable. Once a month we verify that the internal mechanisms of Zerto are working. When I do a test failover we check if VMware tools come up, if the IP addresses change, and the things that Zerto is configured to do automatically. Usually, if there's an issue, it's either I did something wrong when I configured it, or I put in the wrong IP address or the VM itself has an issue, the tools aren't loading correctly or at all, or it was trying to do an upgrade and failed. We've actually been able to identify other issues inside the environment that we would not have realized were an issue by doing these tests.
Our next step is not so much increasing the capacity but protecting things to the cloud. We'd like to be able to take those same 350 machines or so do we have, and definitely, if not the important 50 that we have, but all of them, have them not only go to our disaster recovery site but also split to AWS. It's where we have both of the sites, one in one location and one in a vastly different location and if for some reason, one were to go offline, we would have those objects in AWS to be able to spin up and do what we need to do.
We ramped up from that 50 to 350 within a year and Zerto just took it and kept on running. We are still about the same RPO as we were before, we're protecting 60 plus terabytes of data at this point with those 350. It did what it had to do to create new virtual machines, depending on how many disks there are. I think that I was able to scale with our needs really easily. 73 terabytes are what we're protecting right now across 357 VMs, and we have a seven-second RPO. It went from a small number to a very large number. The issues that we've had around Zerto protection has either been that networking wasn't sufficient, or the storage itself had to be increased.
There are three of us who work with Zerto, that's it. We do contact other teams, often our networking team to get an IP address for something. But when it comes to doing the testing, when it comes to doing the implementation, and when it comes to doing verification processes, it's all my team of three people.
I am the data management supervisor and then I have a lead storage administrator and a senior storage administrator.
Prior to Zerto becoming our disaster recovery product, we were using Dell EMC's Avamar for backup recovery and for disaster recovery, which we quickly realized was not going to work out very well. We used it for about four or five years. When your disaster recovery test is five days and you take one and a half to two days to do restores only, that doesn't leave a lot of time for testing. Now, we're able to do the restore in an hour and a half. Then we actually can start testing the exact same day that we did the restores. In most instances, we're able to actually finish everything within 24 hours.
When we first purchased it, the backup portion did not exist. So having backup and DR in one platform really wasn't that important to us. We use Rubrik for backups and longterm retention at this point. We really don't have any intent on using Zerto for longterm retention, as we're extremely happy with Rubrik. But time will tell if we decide to switch over to the LTR portion of the product.
Compared to Avamar, Zerto is extremely easy to use. I can bring Zerto up and start recovering, failing over, or testing machines before I can even log into Avamar. Avamar was very clunky from its interface. It's very easy for Zerto to go in and recover a machine to a certain point in time. Where moving around in Avamar, since it was Java based, would take quite a long time to get from screen to screen. And the workflow was not user friendly at all.
We have different use cases for Zerto and Rubrik but I think that the interface and functionality, as far as what I get out of that particular product, what its purpose is, they're both about on par. Honestly, we've told both companies before, we would love for one to buy the other so that we can get the best of the disaster recovery with Zerto and the best of backup and recovery, longterm retention type things with Rubrik. Because they definitely are probably the two best products for their market segment.
Replacing Avamar has saved us on the cost needed to manage them. As far as management goes, we still use the same three people. But as far as renewal maintenance costs, definitely. Dell EMC is very proud of their products and their renewal maintenance costs were rather large compared to what we do with Zerto.
Initially, we saved about $1,000,000 three years ago by switching to Zerto. Zerto and Rubrik replaced Avamar. But buying both products together, versus what the renewal/upgrade costs would have cost us for Avamar, with all the hardware, was a savings of $1,000,000.
The initial setup is very straightforward. I built a couple of virtual machines to run the manager on, deployed some VRAs, and then attached it to VMware and checked what I wanted to protect. We probably had it up and running in about an hour total. Then we tested protecting some machines, and we had some test boxes that we tested back and forth. It was a very easy setup. People are definitely sold about how easy it is to install and configure.
Initially, our deployment strategy was to protect a small subset of very important machines for an enterprise. And then once we saw how easy it was to implement, how easy it was to put things in there, and how easy it was to protect them, it went from a handful of machines to 350 or so. The initial intent was to protect a very small number. That went from that to a very large number very quickly. Zerto was able to handle it no problem. We actually had to end up buying more storage on the target side because we had not planned on doing that many machines from the initial implementation.
We worked with our account team. We were able to get the proof of concept software, a link to download it. They gave us a key, they gave us a little Excel sheet stating how many machines and IP addresses we needed. Then they basically sat on the phone with us for the hour with WebEx. And we set it up just that moment. That's really the only implementation help that we've ever gotten from them. Everything else has just been pretty much us on our own.
Their support has been very, very good. We've had some technical issues that we've been able to work through with them. Nothing major, but if I have a question or if we run into an issue, we're able to either open up a support ticket and they respond fairly quickly, or we are able to do some searching in their knowledge base. We've had an instance where we did the upgrade to a new version and it caused some problems. But within, I'd say a few hours, we were able to correct it because they had already experienced that. And they had that logged in their internal database of issues. So, they were able to log in, and give us the fix that we needed and get us back on track.
It definitely is a very robust product. The feature set from 4.0, 4.5 to now has increased greatly. We do like the fact, even though we're not using it, that as long as I pay my maintenance when the new features come out like longterm retention, analytics, the monitoring, the reporting, the things that were not there when we first purchased it that are there now, is all part of maintenance. It's not a bolt-on price. They don't charge extra. It was one of the things with Dell EMC that was always a pain was. They had additional costs. With Zerto it's like, "You paid your maintenance, here's a new feature, enjoy!"
They have licensing breaks as far as 50 users, or 50 VMs, 100 VMs, 250 VMs. We ended up with a bunch of 50 at first, and all of our maintenance renewal dates were all different. It ended up costing us more because we didn't just make the investment up front to say that we wanted 250. We had to end up going back and resetting all of our maintenance dates to the same date. It was just a nightmare for our maintenance renewal person. If you did a proof of concept and you like it, definitely make the license investment upfront. That way, you're not trying to piecemeal it afterwards.
Licensing is all-inclusive, there are no hidden fees.
We looked at RecoverPoint for VMs. A long time ago, one of the companies inside this enterprise had used RecoverPoint and it worked really well when it was the physical RecoverPoint. But as things became more virtual, it no longer was as good as it had been, so they had discontinued it. RecoverPoint for VMs was definitely not as easy to set up. It was not as easy to use. It took a lot more resources. This is three-year-old information, but I feel like we would have had to have had more people on our team than we do now with just the three of us. We didn't feel like it was as stable. It certainly wasn't as easy to use, test, or get to work as Zerto was.
My advice would be to do the proof of concept. They're very willing to help you with the installation. Do a proof of concept. If you're not amazed by it, I would be surprised. Everybody that we've ever talked to about this and have done a test of it says, "I can't believe it's just that easy."
I would rate Zerto a ten out of ten.