This week I have immense, newfound respect for vendors that operate high-reliability networks. A good friend of mine works for a 5 sigma (99.999%) guaranteed uptime B2B integration vendor. She got the “pleasure” last week of receiving the “special” blackberry, which means she was on 24 hour per day call, for 7 days, for any issues the network may have. In her role as second line of support, she had to field calls within 20 minutes of report or the issue was escalated. Her assignment was a normal part of her employer’s sophisticated way of providing first, second and third level support across a network that includes two data centers and primary/secondary servers in both. First line of support is provided a dedicated, 24×7 staff. Second line and third line are staff software engineers, serving in rotation. Some weeks that blackberry goes off very little; other times it goes off a lot.
What could go wrong with a B2B integration network, and especially at 2 AM on a Saturday night, you might ask? Well – there were issues following a network card failure in a back-up server. It wasn’t rebooting properly when repaired. After some sleuth work, the issue turned out to be a bad configuration file. There were missing files from partners with scheduled batches due for arrival. There were rejected files because of duplicate transactions. The list goes on. My friend was one of the less lucky ones – she was answering that blackberry a lot. She said it reminded her of having an infant all over again. You sleep when you can and take care of the crying “baby” when you need to.
No wonder Gartner is observing a shift of B2B integration from behind the firewall to integration as a service and even full outsourcing. 99.99%+ uptime is no trivial task. And the worst part is that that crying baby never grows up!
Category: Uncategorized Tags: