Whew, this was a long week! It’s after 4:00 a.m. on Saturday morning, and I am just about ready to to collapse! I spent the week doing mostly non-technical stuff for the Synchrony Products we’re running. Because of the frequent outages with Integrator, I designed, revised, and finally published “Axway Synchrony” outage checklists for our production and external environments. They’re pretty simple documents. The idea is essentially “If we get production down robots, call me, and call unix admin, and we’ll fix it”.
Since we’re now required to ensure that the product is stable and reliable 99.8% of the time, up .1 percent from last month! What’s tough though is that even such a small increase means that we lost 40 minutes of downtime we could count on last month and prior months that we can’t now. I believe that when all is said and done, we can’t be down for more than 87 minutes for the entire month. We implemented some pretty complex fixes this week in production that are supposed to help us with the outage issues that we’re having, but we’ve had 2 “bliips” already since we added these changes (increasing the number of hierarchal message processors and processing engines, then segregating them by flow type, i.e. routing, ACH, mapping). The blips happen when files are receieved in Gateway, are routed to integrator, but then don’t go anywhere or do anything after that. This normally means we will have an outage, but we got lucky Friday at least!
On a more disappointing note, those GIK errors that we’ve been receiving were supposed to be resolved with a patch that was provided to us, but that we can’t actually apply because it changes the attributes that are passed from Gateway to integrator, which means that we would start to fail on all of our transfers because Integrator and Gateway wouldn’t know what to do with these attributes and how to agree on the communication. As such, my next documenting/process improvement type of project is going to be the monstrous upgrade from Gateway version 6.9 to 6.11. It’s been done in the development environment (1 of 4), but we have a ton to do to stabalize that environment before we can roll out into QA, external, and finally prod. I’ve gotten a good jump on the plan, but it will need to be executed with the utmost precision, so the process documentation that goes into the planning has to be spot on!
The good news is I’ll be busy for quite some time. The bad news is I’ll now be more entrenched in the Synchrony Suite, and I’m starting to learn that the other stuff out there is pretty good stuff, and there are some companies making software that looks a lot better, more efficient, and is ultimately exponentially cheaper than the Axway software. As a consultant, it’s great to be so skilled in the Axway stuff, but when I go to a client who needs to implement a new solution, they’re going to take the one that’s tens or hundreds of of thousands of dollars less, and doesn’t have as many problems, right?
I guess we’ll see, all I know is January is going to be a busy month!
Until next time,
Tony
