What is our primary use case?
It is backing up customers' servers. On the machine that they wish to protect, they deploy an agent on it.
It is essential for all businesses to back up their IT systems. In our view, it has to be automatic, offsite, and require no user intervention at the client level. The SolarWinds product provides all of those things.
The solution supports full-system, bare-metal, file, and folder. One of the reasons that we like the system is that it maintains up-to-date standby virtual machines, which can be booted at short notice if the customer's primary device fails. So, the predominant recovery methodology that we have is backing up to hypervisor as a VM. We have used the bare-metal recovery in the past and that has its use, but all the ones that we have put onto our customers in a commercial sense are all recovering to hypervisors.
Without cloud storage, we wouldn't do it. If it didn't have offsite storage, we would not be using it as a backup solution.
How has it helped my organization?
Because we have confidence in the backup, we feel more free to experiment with our customer systems. For instance, if we have a server, and we've sat there scratching our heads because we really need to make this big change on the server, but if it goes wrong, then the customer will be screaming at us for a month. We can now go, "Oh, it doesn't really matter. Just do it. If it all goes horribly wrong, then 30 minutes later, I'll have pulled the backup from last night. It will all be good.
Because we can keep images of our clients' machines on our test machines in our workshop, I can go, "You know what? I'll spin up their copy from last night, make the changes I want to make, and if it all goes bang, then I'll just delete it." Then, tonight, they'll put a fresh copy on it. So, it's freed us up from the worry of working on live machines.
From the customer's point of view, if it all goes tip top tomorrow due to things like ransomware, they have a backup from Friday. So, who cares?
I have used the solution for complicated recovery scenarios, such as a complex database recovery, to the point where SolarWinds is now using some of the techniques that we have developed in our office as part of their mainstream products. I find it extremely easy to manage the solution in such scenarios, but then I have immersed myself in the product for five years. If you are experienced at building and using virtual machines to get backups, e.g., if you can drive VMware or Microsoft Hyper-V, then I would guess the recovery process is almost trivial. You literally download the software, type in the access codes for the date set that you want, and press "Recover". It is pretty much that simple.
The recovery speed from the cloud depends on how big your data set is, how fast your broadband connection is, and how spry the Internet is feeling that day. Here are a couple of examples:
- We have a remote system in our stockroom downstairs where we can do test builds. If I tell it to build a server that it hasn't built before, it has to pull the data out of the cloud, and and it is sort of a standard size server, then you are looking at about four hours of time. However, we have good broadband, and it's being built onto a pretty potent system.
- In the normal course of events onsite, as well as having a cloud image, it also keeps an image copy on a local data store. In which case, I can pull up a server back for a client in about 15 minutes. This is something that I did the other day.
What is most valuable?
The biggest thing from our point of view is its reliability. It gives us very few problems compared to other solutions that we have trialed.
You can maintain multiple copies of the hot standbys, which is a huge benefit to the protection of customers.
It is adaptable.
The efficiency of the solution’s resource and bandwidth use when it comes to both backup and recovery is extremely good. Most of the time, I'll arrive in the morning, grab my morning coffee, fire up the console, and it's all green ticks, then the job is done and we send the customer their invoice. It's difficult to think of any way that the efficiency could be improved because it all just sort of works in a sensible time.
What needs improvement?
Commercially, they offer the product in two different formats. There is the full imaging backup, and there is also an alternative. You can pay for simple data backups and pay by the gigabyte that is consumed. Unfortunately, you cannot have those two products in the same dashboard. So, I have to switch between dashboards to look at:
- All the servers being imaged.
- All the private laptops who have their "My Documents" folders backed up.
That is a bit of a hassle, but it is not a deal breaker. It would be very nice if it was all on the same dashboard. I check our clients for the imaging product (the expensive one) every morning. I check the people who are paying us for data-only backup once a week. Therefore, once a week, I have to log out of portal A and log onto portal B to check if it's all good, then I log back onto portal A. It would be nice if I didn't have to do that, but it's certainly not something that keeps me awake at night.
We don't use the solution’s automated recovery testing because SolarWinds made me cross. When they released it, I went, "Oh, well, that's quite good." Because if you use the system, then it supposedly spins up, and on the portal, it gives you a screenshot of the booted device. So, I phoned up, and I said, "Oh, that's really quite cool. How much is that?" They said, "No, no, no. It's all included in your license." I went, "Okay then," and went and deployed it on about half the fleet. One of the options that our customers have is they can pay us a small amount every month for us to test the recovery just to prove that it's viable, and I thought, "Well, this will do that for us. Nice." Then, in the next invoice, we got a charge for it. While It was not a huge amount, I took offense at the fact that we were told that it would be a no extra cost option that was part of our license, but it turns out that it's chargeable. Therefore, we haven't used it since.
For how long have I used the solution?
We have been using it for about five years.
What do I think about the stability of the solution?
I don't think that we have actually had any downtime since we started using it.
What do I think about the scalability of the solution?
As far as I can see, there is no limit to its scalability.
We have 13 servers and about a dozen private individuals on the data-only version.
How are customer service and technical support?
That technical support is excellent and knowledgeable. We haven't yet thrown anything out to them that they haven't been able to fix in relatively short order. Of all the IT companies that we deal with as suppliers, I'm pretty sure they are the best that we have ever dealt with.
It is very seldom that I throw a question at them where they have to punt it up to the next level. So, their Tier 1 support is really first class. On the odd occasion that we have had to go to their Tier 2 support, those are the developers of the product. Therefore, you get straight through to the developers in Tier 2, which is very good.
I have a problem at the moment, which is probably down to the definition of the host machine, not the backup solution. It seems to be something integral to the VMware version that we're using on a specific machine, but we've ended up, after two or three conversations, with a situation that works. Even though it's not particularly elegant, it meets all my criteria, it's automatic, and the end result is recoverable. The fact that it puts a warning up once a day on my dashboard is something I'll just have to live with until they source it out permanently. However, it's the exception rather than the rule.
Which solution did I use previously and why did I switch?
The solution that we had in place for this tier of business prior to us doing SolarWinds was a limitless nightmare. I reckon that I spent probably 75 percent of my working hours managing the system, and it was not working properly. It was truly hideous. That has now gone down to 10 minutes a day, which is just me checking all the backups. If there is a warning, I just check that the warning is something that I can live with or determine if I have to take a bit of remedial action. Therefore, that percentage has dropped dramatically. However, it's been so successful that we now have a lot more customers using it than before. Obviously, that means the work load goes up a bit, but that's fine. That is what we're paid for, because more customers means revenue.
I struggled with our previous solution for 18 months. It was practically my full-time job for 18 months. At no point, in those 18 months, did it do a backup that was restorable.
We used the Max Backup product in a very small way for a couple of customer's laptops, so we thought that was what it was. SolarWinds invited us along to a little seminar locally in Birmingham. We went along, and a simple, "Yeah, we quite appreciate the fact that you're buying 20 quid worth of backup offers for these two guys, but you do know we've got enterprise class products as well?" They made a very good presentation and impression on us. They discussed the pricing, and it all seemed very advantageous, so we signed up the following day.
How was the initial setup?
The initial setup is ridiculously simple, but then again, I've done so many of them. You set it up on the site and that gives you a device name and password. Then, you download it on the client, install it, and put in the device, then put in the password. After that, you set up the encryption key and press "Backup". It is not hard. The killer is if someone loses the encryption key, then the whole thing is a waste of time.
Because you can only recover if you have the encryption key, we are very careful in logging and maintaining the encryption keys for our clients' backup systems. Though, I can just imagine someone who writes it on the back of an envelope that his tax demand came in and that accidentally goes in the skip. Then, three weeks later, when he needs to do a recovery, he can't find it and that will be a problem. That is the only "Gotcha." You have to be very careful with your credentials for the backup.
The deployment takes me longer than anyone else because I understand what I am dealing with and take more time and care.
- Takes 30 seconds on the platform to generate the new instance.
- Takes a couple of minutes to download it on the client, then a couple of minutes to install it. So, it takes five minutes get it installed.
- I will then take half an hour just testing the various scenarios before I turn on the scheduler and let it get on with its own stuff.
If you are spending an hour doing the deployment, then you are doing something wrong.
Once it is installed, I just choose the smallest file that I can find and back that up. I check that the backup goes through okay. If it appears in the local cache, appears on the web platform, and I can copy it back to location of my choice (all of which takes five minutes), only then do I know that the backup is fundamentally sound, and go, "Right." Because the thing with incremental backups, the first backup can sometimes take a week because you're backing up absolutely everything: Gigabytes and gigabytes of data. It's only after that first backup is done that it becomes incremental, then it does it in a couple of minutes. So, I'm not going to unleash it to do a backup that might take 30 hours only to find that I have had a bit of trouble with one of the settings, and it hasn't gone how I thought it was going to go. So, my implementation strategy: You just back up a file, check that everything works exactly as you expect, and then let it get on with it.
What about the implementation team?
I do the deployment and maintenance. It takes an hour to set up a new machine, then I spend 10 minutes a day checking it.
if you want to phone up a reseller, like us, we don't put a big margin on it and can make it work for you from day one. Whereas, if you go direct, you will trip over a few things, which we tripped over five years ago. There is no need to trip over them again and reinvent the wheel. For example, when you're working in an enterprise environment inside Active Directory, you set up the machine which will host your local copy. You need to authorise the share to the backup system. You set up a share based on the Active Directory username that you have created to manage the backup system, and it's all good. If the machine that you're protecting, i.e., the Active Directory controller, goes and fails, then you cannot get to the speed vaults (local copy) because it's protected and only accessible through local directory users. The answer to this is that you set up a local user inside the machine which hosts the copy, accessing using the local user credentials, and then it's all fine. However, that is not obvious, and it's not documented. You only find out about it when it all goes horribly wrong one day, and you can't get to the local speed vault.
What was our ROI?
We have absolutely seen a healthy ROI.
Something that we make a little money on is if a customer wants to pay for the option to have an onsite copy on their premises (ready to run). We then supply the hardware for that. It tends to be a fairly low spec server. We either sell that, in which case we make a small margin on it, or it goes in as a rental cost on the monthly.
The solution has reduced backup times by an immeasurable amount. Its backups are incremental, so you are only backing up data changes based on the last 24 hours or so. If you are also maintaining the stored images, the restores are also only incremental, happening in minutes. Whereas, with a lot of the other solutions that we have looked at, each time it goes to refresh the restore, then it has to build a completely new image. That takes forever. This solution also improves recovery time.
SolarWind's total cost of ownership is an extremely good value.
What's my experience with pricing, setup cost, and licensing?
There is a basic product and pricing package. Clients can add on additional pricing for the local storage and hot standbys (onsite and offsite). We like that we can build a solution that fits the client's requirements and pockets.
When you have a device/appliance on the site quite a lot of the other enterprise class backup people insist that you have their appliance, which is frankly offensively expensive. Because when you pry the top off, it's just a standard 19-inch tin box with a standard Intel I5 in it, some RAM, and a hard drive. Then, you go, "Why have they just charged me 5,500 quid for a box, which I could have probably build for under 500." Whereas, with the SolarWinds product, they don't have that. The backup appliances that we have onsite are just plain cooking PCs. We can build our own machines, which is reflected in the price that we can offer a customer.
There is something you have to do each time you effectively buy an imaging/server license, which is a fixed price. So, if I set a new machine up on the portal to be backed up, it will cost us one more license and appear on our next invoice. With each license comes 500 gigs of cloud storage, which is pretty much as you'd expect. The nice thing about the SolarWinds product is you pool it. Therefore, if you have 10 devices, you have five terabytes of storage, then we can divvy that up however we like. For example, if we have one customer who has a tiny little machine that has only 200 gig and another customer who has a machine that is 700 gig, I still only need two licenses. This is because the 700 gig and 200 gig make 900 gig, which is less than the one terabyte that those two licenses give me.
This pooling means it is very cost-effective from our point of view. It is just an example of where they have built the system to reflect the customers' needs not to maximise their profit. Because they could quite easily say, "Nope. As soon as an individual machine hits that five terabyte limit, you have to pay the sliding data scale to have extra storage on it." They don't, they go, "Well, it's gone over its limit, but you still got stuff left over from some of the other machines. You can use that for free." That is a very good indicator of a company who is customer-focused.
Which other solutions did I evaluate?
We have probably evaluated all of the solutions in this market. Technically, they all do pretty much the same thing. Their core product, which is automatic backup, goes into the cloud and leverages the VSS inside the server.
SolarWinds is the only one whose even close to that level of flexibility and modularity that we need. There are many offerings in the market, but the reason we chose SolarWinds is that it is modular. For each individual client, we can put a system together with as many features as they want, need, or are willing to pay for. In other words, we can go from a very minor backup operation for only a few pounds a month to many hundreds of pounds for the full solution.
With due respect to various persons around the planet, if I have a backup system with a customer who's screaming at me because his server has gone down, and he wants his backup running, then I don't want to phone up and be answered by a call centre in the Philippines telling me that someone will get back to me, because they're in some far-flung parts of the world and the call won't come through until two o'clock in the morning. That's not going to float my boat.
When I phone the guys up at SolarWinds, I am on first name terms with all of them. Nine times out of 10, I will get through to one of the guys inside of 30 seconds. Normally, within 15 minutes, the problem is resolved. That is worth more than money.
Those are the two things differentiate them: the flexibility and modularity of the system along with the quality of their support. At this level, cost is not the number one driving factor.
We use one other backup product. It is considerably more expensive than the SolarWinds product. However, what it does, it has integrated into that product cloud virtualization in the event of the customer losing their premises as well as their hardware. In other words, their backup systems have gone as well. For example, Worcestershire, England spends three months of the year under water. We know it's going to happen, and we just get on with it. However, if that premises was flooded and they lost that backup as well, they can all decamp to an office space, somewhere local. Then, we can spin the whole infrastructure in the cloud, where they can VPN to the cloud, and within days, carry on as normal while we rebuild their infrastructure.
SolarWinds does have something similar, but unless you're willing to pay for a vast amount of storage, AWS is too expensive to maintain. For example, if the flooding scenario happens, you have to provision some AWS space, recover into the AWS space, and then you can do it. You are looking at 24 hours, maybe more, to make that happen. Whereas, the other product, it's literally press the button that says, "Virtualizing Cloud", and by the time your cup of coffee has got down to a drinkable temperature, it's all working. That is the one feature that SolarWinds is missing. Having said that, I wouldn't want it, if it puts the cost up significantly, because most of our customers do not consider that eventuality sufficiently likely to be worth paying for.
What other advice do I have?
It is a very good system. We have very few problems with it.
Just do it. If you get stuck, find the guys in support and they'll talk you through it because it's very quick and easy. They are quite happy to come onto a Remote Desktop Session, and go, "No, you've got that wrong."
It's a good product that is sensibly priced. Anyone who has a modicum of IT skills can make it work.
I would rate the solution as a 10 out of 10.