 We're starting a monthly series here called TFIR topic of the month or T3M. The idea of the series is to take a pulse of the ecosystem of the industry and then bring together experts to deep dive into these topics. I sit down with C-level executives, analysts, engineers and experts to explore these topics and then bring these discussions to you. Today we have with us Eddie Smith, Solutions Architect from Scios Technology. Eddie, it's great to have you on the show. Hi, thank you Swap. It's so great to have myself on here as well. And yeah, really looking forward to speaking to you. What kind of trends are you seeing in context of cost cutting or companies becoming more cost efficient? So I think the big one is a lot of companies now are moving into cloud space. A huge amount of them now are trying to leverage the cost savings, the efficiencies, and the ability just to switch on and switch off as neat as need to be. That saves costs, but I am seeing some companies now they're looking at moving back on-prem. One of the main reasons being is there are some hidden costs for cloud that they maybe just didn't appreciate. Maybe they're not quite disciplined in switching things off when they should do. Maybe they've refreshed their internal hardware to a point where they say we don't need the cloud anymore, whereas when the environment was incredibly tough, someone coming up to them and saying let's have a multimillion pound refresh if you'll fight your old systems. It was just so much easier to go to the cloud. So I'd say that's probably the big one. And then the one that's always in that is the advancing march of the software companies. So you're getting virtualization that's now been taken over by cloud, as I mentioned, but you're also getting a lot of software-defined stuff coming in as well. So networking being done at the software level that can change cost efficiencies will certainly improve them. A lot of the software storage under the operating system level is more complex and a lot more cost efficient in terms of the way that it does various things. So I can think straight away at the top of my head, the storage layer, you've got things like deduplication, you've got compression, you've got compaction. And all of those things can reduce your storage footprint whilst also removing the amount of space you need in your DR site as well. So you need all of that in your DR site if you were to maintain a highly available solution. When we look at cost cutting or cost efficient, there are a lot of things that you have to cut, but there are a lot of things, there are a lot of practices that actually make you cost efficient. For example, there's nothing worse than downtime because it's bad for the image of company as well, and you also lose business. So there are a lot of practices that actually help becoming cost efficient. Talk about the rule of HA in making companies actually more cost efficient. HA is incredibly important for pretty much all the customers I deal with. They're very likely to be in a sector where an outage is going to cause a huge problem for them in terms of reputation loss. If they're a public facing company and something breaks, you see it quicker than Facebook, no, probably if something like that goes down. Public facing cloud solutions, if something happens there, that's obviously reputational for them as well. And then in finance retail, if anything happens in those sectors, you've got monetary value associated with that loss as well. So your online systems are not able to trade or they're not able to sell stuff and shop, that impacts the business and their underlying financial figures at the end of that quarter are going to be as good as they should be. You've got industries like transportation, energy and utilities, health care. Another big one is the health care system goes down. Some extremely important life changing technologies used within health care. So an awful lot of these companies and these industry sectors are actually deploying HA solutions because they recognize that to deploy a HA solution it minimizes the downtime, which is their most important thing that they want to minimize. And it reduces complexity than getting a human inborn and saying, hey, what's going on here? Let's get an engineer to site. Sorry, it's going to take you eight hours for the guy to turn up and fix where it went wrong with the hardware before we can switch those systems back on again. A HA solution will actually automate all that for you. It'll take all your systems, move them to a secure DR site in a second location away from where the problem occurred and it will allow that HA system in its standby state to come up and continue to service the application or the services that you lost in the original site. So most companies that I work with, in fact, there's not many I've seen without it, have a HA solution in there somewhere. And that's the way customers deal with it. Since you touched upon some industries, I want to go a bit deeper into that as you folks work in that area is that are there any specific industry sectors which are particularly concerned with cost and efficiency savings for their HA environments? Financial services, they do require high levels of availability to avoid the down time of potential losses, but they're also incredibly interesting saving costs as well. At the moment, we're seeing quite a lot of layoffs in the financial industry. They're reducing headcount as their primary method to save costs and they're also relying on automation as well. And that's not just true for financial services. A lot of customers are actually going down the automation route as well as the cloud route but backed up with HA highly available systems at the same time. And I think that the thing to say is that all of these types of technologies are great and they will save costs and they will save money because it does avoid those outages that cause financial loss to particularly people in the financial sector. Healthcare, it's always been there to be honest with you. They've always had some form of resiliency in there to protect it because they're critical. If something goes down there in healthcare then people die. In terms of cost efficiency again, it's staff, it's renewal of technology. It's been incredibly competitive with their vendors to say hey, we can't afford to pay either a refresh this year so we're going to pay increased support costs but overall the environment actually costs a little less. And things like the pandemic and the way that the global economy at the moment is just looking is a lot of people are pressing as many places as they can to try and reduce those costs. The finance service has been one that I can immediately see government recently. I know in the UK there's lots of people out on strike at the moment because they all want pay rises and the government are not allowing them to have a pay rise or not as much as they want. So again it affects people individually when these companies want to go down the route of saving costs. But they're the main ones I see. We are talking about cost efficiency, cost cutting and also HAA. Can you also talk about when we look at HAA solutions, how do they in a way help in achieving? You did touch upon that but I want to go just specifically in details of that as well that how they actually help these companies because these companies as you said they are very very sensitive so they do have all the practices when they come to high availability. So let's talk about what they're doing so as to ensure that they don't compromise on high availability solutions but they actually help in achieving cost efficiency and cost saving. So there's a broad number of categories here where they tend to achieve these savings. Some of them don't immediately show that they're going to save you costs but they do. So the easy ones that everyone does monitoring and alerting. It monitors all of their systems. It tells them when something's gone wrong. It could be an email. It could be an SNMP track. It could be as simple as someone just walking around the data center and noting down the lights on a piece of hardware. But they go through monitoring and alerting processes internally to try and get a bit of an idea. Some places are going to the level of doing proactive health fixes as well at the moment so they can proactively use an artificial intelligence and environments like that to say hey we've seen this it may not be an error and it may not cause an outage yet but we've seen something here that we think if it does it three, four, five times you've got a 50, 75% chance extra of the whole thing going down. So you can do proactive health care based on the output from your monitoring and alerting systems. And sort of finding out when that problem is going to happen before it happens. The next one is capacity planning. So again I mentioned the storage world. I've been to so many places where core capacity planning has caused an outage. It's caused the system to go down. When all it would have taken is for someone to go look at that system and say hey let's clear up some of the monster stuff there, stuff that we don't need anymore, stuff that's been there for years and nobody's touched. Let's get that off onto a backup system or let's just delete it, get rid of it or move it on to a lower tier storage that's less expensive than the one that's running their primary services. If your production database goes down because some users have stored MP3 files on there your business manager isn't going to be very happy with you. So effective capacity planning is quite an important one as well. So one that I've seen myself is companies ignoring maintenance and performing upgrades in a timely manner and this can be as simple as applying the latest hot fixes or patches onto your operating system or performing operating system upgrades. I think a big thing in the world at the moment is cyber. So the lack of maintenance, the lack of upgrades can expose you to potential security holes that could take out your systems cause reputation or loss. After you've had embarrassing questions with customers that you may not want to have, get legal people involved and all manner of things just because somebody couldn't proactively go in there and say hey we need to upgrade the system this weekend can you give me a couple of hours to do it and you close off those holes. And it's also just good practice as well. Keep your systems in top shape, follow the advice that the vendors give you and for the most part you're going to get good support. Whereas if the first thing you do with the support person is say I've got to touch my system for five years they're going to say well hey go apply the latest hot fixes and come back to us when you have and if you still have a problem we'll help you. Another thing that is one of the things that gets done in the background and nobody really appreciates is your backups. So these things just happen. They go on, nobody sees them but when something goes wrong they're the first place you're going to go to pick up the phone to your backup guy and go I've just lost my production database or a file after spending two weeks on it and I've now lost because I've accidentally deleted it or it's got corrupted or something has happened please Mr Backup Man can you help me. And it's great when you get that return back going yeah when do you want it, it's going to take me 15 minutes and we can go back to last night. It's a worse case when you ring up and you say oh we've not done a backup for two years. Or oh yeah we thought it was working but actually it's not actually worked for six months but nobody noticed. Which is where the whole thing around monitoring and alerting comes in. You know you should pick up that stuff as it happens and go and fix it before you get to that stage. But backup bands and recovery techniques within an organisation are extremely good at trying to maintain uptime because they get you back to a certain place that you know in a relatively short amount of time but it still does take time. It's certainly not as good as a full fledged HA solution where it's pretty immediate when the failover occurs. So another way that customers and companies are actually reducing outages and increasing cost efficiencies is through the use of redundant systems. So this ensures that critical systems can continue to operate even if one component fails. In partnership with that critical systems as well as having multiple items that are redundant they'll have an HA solution in there as well. So not only do you have protection from a component failure you've got something looking at that piece of hardware and telling it to go and failover to a standby site because you know hey this component's gone it's going to cause a problem let's failover to a different site and then allow the engineer to go in and fix or the software engineer to go in and fix whatever it was that went wrong with that particular component. And I keep mentioning hardware but you can get software failures as well so it isn't just pieces of equipment that go wrong it can be software as well. So we've mentioned it briefly but disaster recovery. So disaster recovery is where you pretty much lose your entire site. That's the way that most people think about it what happens if this data center just for some reason disappears where do we go from there and continue to operate. A lot of companies operate a DR strategy where they say okay if we lose our main data center we're going to switch over to the other one. And as part of that we'll have a set of redundant systems over there that wouldn't be running in normal situations but in the event of a DR scenario we can bring them up. An important component of that is making sure that the data is over there as well because it's no good bringing up a bunch of systems in a DR site if the data that they're attached to is out of date. So that's complimented usually by some form of data replication between the two sites and that can be provided at the host level so using things like volume managers to do that for you or it could be done at the storage level so the storage vendor if you have one can put their own technology in place it's good for the storage vendor because they get to sell twice the storage and it's good for the vendor because they get the licensing costs associated with it but it's also you get the full support and you get the help and advice and professional services and other cases where you need to help setting that up or actually implementing a DR plan in that worst case scenario where something does go wrong. We have HA solutions that help with this as well so we've got some HA solutions actually our DR capabilities built into the product so they will do not only the failure of the services from one site to another but they'll also do under the covers they'll do the data replication for you so not only does the application or the database go across but the last good update from the data which could be literally it up to date to the primary that can come up in the DR site as well. The other one that I do see quite a lot of again in the cloud environments particularly is load balancing so you'll put your client in front of a load balancer and your load balancer will have health probes in there that detects whether multiple systems multiple web servers are actually operating and then direct the traffic in the most appropriate place and they've got clever algorithms that can detect which in my example a web server which one's busy which one's taking time to service the requests a little bit slower than the others and just not send data over that way because they know we're going to get a slow response or a health probe that will detect an entire piece of hardware has failed behind the load balancer and actually switch it to something that still remains alive and that's another way of doing a check as well so there's quite a lot of different technologies in there to try and keep things going but as cost-efficiently as you possibly can. So guys, do you have for companies or teams what a step they should take to ensure a good balance between cost-cutting, cost-efficient and also having a very well-defined disaster recovery and high availability? Whenever I go to see a customer and they say we want to achieve DR solution or HA solution for both but we don't want to spend a lot of money and it's a difficult conversation to have because you know straight away it's going to cost them more than it will do for the most part than just having a single site with a single node if that goes down or the node goes down they're faced with in some cases lifting and shifting that box in the back of why go over to another site that has the network, it has the infrastructure it has the connectivity to your clients it's a case where you say look this is going to cost you and it's just a matter of how much it's going to cost you and I'll tell you some of these earlier your data is usually the most important thing but you don't want to lose so in the event of a data corruption in the event of being hit by a malware virus which encrypts your data or anything that affects your data you want to protect your data is the foremost thing because you may not have something to serve the data out but you have it there and it could be sensitive data as well so you could have customer information in military and governmental type of parliaments it could be sensitive politically or it could be important from a depending on which agency you're working for you don't want that sort of information to get out to anywhere else so you want your data to be clear so your first thing that I try to say is look put your data first and then work out how to protect it work out how much that solution is going to cost you is it going to be your host based thing which is free generally you've got an rsync, you've got robocopy they're free, you can put them on there put a schedule in place copy the data across a regular interval and you still need somewhere to copy it to so you've still got to buy the storage but you could complement that with some things like compression which is free and all you could go down the additional route of putting some expensive sound storage in there getting a storage vendor putting in an expensive network that's dedicated just to storage whether that's a fiber channel a fiber channel over ethernet or just a regular iSCSI using a normal network you could do that as well but if you want guaranteed performance you're going to need a dedicated network and some people just don't have the money for that but you do get the value of getting all the vendor support and plus all the features of functionality that they'll tend to provide as I mentioned you can patch, you can compression you deduce your step chops all the stuff that come along with it but they all cost money the next thing that I tend to look at is okay what do you want to do with a host you've got this expensive 16 core multiple gigabyte database that's running on this system here and you want to take it over to your DR site or you want to make it highly available but you don't have the funds you don't have the the means, the cost available to you in your budget that year to just go and mimic that in a DR site or even a second node added to the cluster so this is where some people decide to go down the cloud route and they'll set up a DR environment or a highly available environment in the cloud which costs less there's no doubt about that as long as you only use it when you need it you've still then got the problem where you still need the other data there but that can be done in different ways and the cost of just having the storage in the cloud is definitely not as much as having the storage and the compute in the cloud as well and some people would go and hold they'll say okay we'll put both environments in the cloud and we'll switch on the system when we need it so my working hours are only nine to five what's the point in having these systems running overnight when it's just costing this compute money in the meantime so that's another way that people can save cost is by being very disciplined and switching off systems they don't use when they do it which is obviously really easy in a cloud environment you just click on it and say turn it off in a real world environment with physical on-prem stuff you've got to have people going around switching these things on and they may not switch them back on because you've got moving parts in there like hard drives that fail after so long particularly when you switch them off and on again they don't like it so there's lots of different ways of achieving cost the cases I try to show to customers is you know what have other people done in your type of environment in a similar situation to you guys the unfortunate message is that it will cost you at some point to have this duplicate environment because they're just in case something goes wrong with the old one but you may not lose as much money because you're still able to generate business, you're still able to build your customers you're still able to maintain your financial transactions to the external clients that you use so there's definite benefits and you don't lose that by having a standby system in place but what it does mean is that it's just a case of do you want a Rolls Royce or do you want a Ford Escort that's the difference between the two in this case picking the one that suits you best for the cost that you could afford Eddie thank you so much for taking time out today and I talk about this topic today and I would love to have you on the show back again, thank you No, thank you very much, thanks a lot