Charles Kim, CEO of Viscosity describes a zero downtime migration

Thanks everybody for coming. Hey, my name is Charles Kim. I'm the CEO and founder of a company called Viscosity North America. We are a SharePlex Platinum partner. And I've been doing SharePlex since 1999. And so I worked for a software development company, and back then, we used SharePlex for HA back then as well as for our reporting databases. And so if you look at back in 1999, there weren't many technologies that enabled you to do HA. And so we were able to do disaster recovery with SharePlex back then, and we did it for all of our customers. And since then, I've been promoting SharePlex at numerous companies. We, back in the mid-2000s, brought in SharePlex for a major insurance processing claim company. And we were able to do 30, 40 million rows a day in some of our largest tables replicating and doing claims processing for insurance companies. And so this is for companies like Aetna and Blue Cross, those kinds of companies too. And so since then, I've brought SharePlex in numerous companies, and in particular to one of the largest mortgage companies in the country. And so at this company, we upgraded, at that time, from 10.2 to 11.2 and also migrated from AIX to Linux along the way. And we did the entire migration and upgrade with zero downtime. And so you guys all know what GoldenGate is, right? And so SharePlex has a very high performing, complementary data replication which is just as good as Golden Gate. But the key thing with SharePlex is that you, as database administrators or architects-- the learning curve is almost nothing. And so literally, you can go from knowing nothing about SharePlex to becoming a SharePlex expert in a matter of just weeks. And so a lot of our guys, a lot of my consultants, go through the SharePlex training, and within a couple of weeks, they can do an enormous amount of SharePlex consulting. And so we, just in the past 12 months, have done 20 SharePlex implementations across the country. And including that, what we've done is, for some of these larger distribution companies, we've set up multi-master database replication. And if you look at a lot of your CIOs, they want what you call active-active active data centers. Have you guys heard of that concept? The ability to be able to generate workloads between two different data centers and be able to scale across different data centers. And we're able to do that with one of the largest distribution companies in the world. And we're able to set up conflict resolutions. Whether it be site-based, or whether it be time-based, those come natively with SharePlex. So I work for another company which is a credit card processing company, where we had to basically provide five 9's to our customers. Five 9's. 99.999% uptime. You're talking about literally about 26 seconds of downtime a month. And think about that kind of downtime that you have to do. Imagine adding a column to a table. If that table has, let's say, six billion rows, and you want to add a column, and you want to make that column not null-- exactly, right? But leveraging SharePlex, I can do that kind of stuff. Basically, what you would do as a DBA, an architect, you would actually go write a little robot. A little spider will go an update hundreds of thousands of rows at a time. And eventually, behind the scenes, you will update all the no values and eventually enable the dot null constraint. And so I've had companies where we do three-way multi-master replications. And especially if you're dealing with financial customers, they have regulatory requirements to do switchovers every quarter. And so these guys have to switch over and be able to show to their customers that they can actually not only switch over, but all the data's there. And if you're looking at five 9's-- customers who have that 99.999% uptime-- if the application's not responsive within 15 seconds, they have to switch over. And that's where SharePlex comes in. It's not about just being able to do upgrades and migrations, which all of us can use SharePlex for. But it's also being able to do things like active-active datasets. And so another use case that we do for a lot of our customers is reporting databases. We can actually offload production workloads from your primary database. Instantiate a reporting database that's synchronized almost real time, and you can offload all your reports to the reporting database. Another big use case is if you're going to go talk from Oracle to SQL, and vice versa-- SQL to Oracle as well. SharePlex now can do heterogeneous database replication between multiple databases. So we have customers that we're doing that for. And if you're looking at SQL server customers right now-- all of you guys are running Oracle. However, most companies out there have small pockets of SQL server databases. So if you look at Oracle customers that also have SQL server, you're addressing probably 90% of the customer base out there. And you've got to replicate the data in between. So one of the biggest things that SharePlex offers is the ability to compare records between source and target. And that's one of the most powerful things about SharePlex. When you do data replication, there's always concerns about, is my data replicating to the target database? And is my data in sync? But with the compare and repair capability that SharePlex offers natively, I can compare all the rows. And if there is a deviance, with one command, I can simply repair that data. I can say, repair the target, or repair the source, depending on where the source of truth is. So with that in mind, we've done zero downtime migrations and upgrades. But think about incorporating what we call a reverse replication. So now I have instantiated a 11.2 to