On the morning meeting with Steve, Alan, Wayne, and Dustin. Checking up and getting caught up with what happened over the weekend.

Notes from Wayne about server stability and processes

Infrastructure - We have greatly increased this portion. We would like to fix how we deploy code to the servers. Currently we are just using FTP and are able to push code as needed to different servers.

Code - The last couple server issues have been code related (bugs, poorly written code and queries, or loose code). These code problems are currently affecting our servers as much or more than the infrastructure process.

See attached for a small proposal from Wayne about how to configure our environment. Wayne would like to look into Lucee vs ColdFusion. I put tons of notes on that physical PDF upload.

We talked about some costs and also options going forward. We talked about DNS (domain name servers - where are things pointed), SSL's (secure socket layers - https stuff), emails and texts, etc. We can virtually push some of the logging into the AWS CloudWatch services. Amazon also has some other servers that we could harness and/or use. Some of the other services are AWS Cost Explorer, AWS Budgets, AWS Cost and Usage Report, and Amazon Route 53.

Steve was asking questions about how we could make an adilas user group (power users - adilas community) that could be linked and/or joined to different companies or different adilas clients. Almost an extension of the adilas reps and consultants - who wants my skills, I have a power user account, and that could be tied in to anyone's account (based on permission and access). All secure, but also very mobile. Basically a way to separate users and clients.

We also talked about splitting up the databases (world building) into corp-specific databases (smaller single databases) and then working on the mixed or cross-corp queries. In the background it could be tons of different services and even computers, but in the frontend, the presentation seems singular and very streamlined and smooth. This also opens up more revenue options for our developers to help code cross corp or consolidated reports and such.

Some talks about future white labeling options and how we could setup adilas as a platform and then allow them to pay the AWS bills and they pay us for usage of the code.

The rest of the talks circled back around to the code and how we create and deploy our code. Lots of talks about automated testing and getting all of that stuff super stable. Being able to up-scale to handle bigger and bigger loads. Alan popped in and also brought in the possible option of down scaling as well. No one likes to talk about this, but we need to think about both side of the coin. If we are on a dynamic environment, we could virtually scale up and/or down. Once again, it comes back to a two-part puzzle... you have both code and infrastructure.

Alan was talking about how to encourage our developers to run more of a test driven code structure. This will be a cultural switch. If we make it so that the system becomes the bad guy, we could help to change that culture. We've done this before, when we first introduced permissions and such. We talked about training, tooling, and guidelines to help some of these developers. Most of the talks today have dealt with changes to infrastructure - we also know that there is possibly a bigger conversation about the code development side of things.

Where are we going and/or heading? Eventually, some of the adilas functions and features will become components and virtual standalone pieces (modules of sorts). We need the flexibility of a Legos type scenario... multi interlocking blocks that could be mixed and blended as needed.

From Steve - I'm seeing lots of our wish list boxes being checked with these new proposed changes.

From Alan - This seems great for scaling. Hardware is one of the hardest ways to scale. If we could turn it into a service, it becomes easier to manage. Alan was also talking about scaling up and scaling down. Make sure you can go in both directions. How much weight are forced to carry... being able to absorb or extend as needed.

We can make the whole presentation more seamless. Currently, we send clients to data 10, data 11, etc. If we go more in this direction, it just runs more seamless and we scale things in the background as needed. In the background, we almost need a platform babysitter that helps us know what is going on and/or what is available.

Our clients tend to like - our pricing, our possible functionality options, and how easily we can customize things.

Dealing with timelines... We are seeing this transition between 6 months to a year. Maybe even more. We will start on it immediately, but it still may take some time to roll these things out. Also, we can only see so deep, there may be some unknowns that hit us as we get into things. Alan was saying some of the fun development sayings such as the last 10% takes 90% of the time and other things like that. It always gets crazy. Slow and steady wins the race. Lots of good things on the horizon.