Talking with Wayne, Bryan, and Steve about new servers, specs, and the configuration process. Wayne was talking about a virtual recipe for the servers and then using a program called Puppet that helps with special configurations and even bulk server setups. Bryan has been helping us with some server setups. We are trying to get things transferred over to Wayne and let him run with things. He really wants to do that kind of stuff. He used to do it for his other company that he worked for.

- Looking at server configuration options and such - lots of ideas and options.

- Steve had some questions about inbound and outbound API socket access (external traffic above and beyond what we know about and/or control). This could be outbound pieces to state compliance tracking systems and/or inbound ecommerce requests or inbound API socket calls (general data stuff). Lots of unknown potential and/or uncontrolled amounts of traffic (in/out). That is a big variable and is widely unknown. Steve doesn't think we have even remotely seen what is possible and/or on the horizon.

- Wayne was talking about a frontend software application called Cloud Flair or Cloud Flare. It helps to filter the traffic good/bad and what actually hits the servers. We were talking about protecting servers against attacks and what not.

- We were talking about different models where we break up the databases and keep things all on their own servers. Or we could build super structures and configure a super Linux box with a non web facing database server. We could then have a number of smaller web servers (code files) that connect to the super structure. See the attached for a couple options that we were drawing (possible configurations).

- We were talking about amounts of time, costs, and resources to duplicate things and spin things up and then manage them. Currently, we are on a two-week to a month spin up process for a new dedicated box. We would love to get that down to under 24 hours and/or something similar.

- Steve was talking about doing some costs analysis, prototyping, and calculated experimentation. Lots of good ideas but still lots of unknowns. We need to be smart as we venture forth. We need to keep the main adilas ship rolling along so that we can fund some of these new experimentations.

- As we improve the process, we are planning on circling back around and bringing the older servers and older software versions up to speed. We may end up cycling through the servers and slowly bringing them up to a new standard, as that get established. Lots of new ground.

- As part of the meeting, Wayne was showing Steve and I some of the Nagios (monitoring server) and its stats. We were seeing what updates ran and did not run, what CPU usage was going on per server, and what database traffic load was going on per server. Some of the servers, are close to 2 million queries since the nightly reboot at 1 am. It is only 10 or 11 am and already they are close to 2 million queries into the day. That ends up being between 30-50 (ish) queries per second as an average. That is tons of traffic. With some of Wayne's tools, we can now see some of the stats and trends. Pretty crazy.