Comment Re:Complexity, my dear Watson (Score 1) 1533
Wow, take of the nostalgic rose covered glasses and wake up. I punched cards on and IBM mainframe and let me tell you it wasn't anything like you described. The only highly trained engineer was the IBM service rep. The university I went to attracted lots of geeks to mainfram programming they just didn't have degrees in computer science or E.E.
I'll grant you we didn't have to IPL the system every day and we did average 99% uptime during prime shift. On graveyard shift when the systems programmer make their changes it was another story but since the system wasn't officially available during those times we didn't have count the downtime in our stats.
No, there are only 3 differances that I can think of that really make differance
1) A Support contract - We pay big bucks and when there is real trouble truly amazing things will happen. They almost were going to rent a plane just to fly a part to us once because we were down with a hardware failure and we were approaching the 24 hour limit in our support contract.
2) Strict control over hardware and to a lesser extent the software on the mainframe. If some issue of hardware or software arose and fault was serious enough IBM could void the support contract unless it was fixed. In fact a bug was was found once in the OS and IBM contacted us and said that you must install this patch to to the OS or you will no longer be under support.
3) Finally, the complexity and 24/7 nature of computing today has really stressed machines far more than we used to back in the good old days. While we didn't IPL every day we did "roll the regions" every night starting at 4:00 AM just to make sure things would run smoothly the next day. Now we can't do that because we've got a region that needs to be up 24/7 to support the web.
I'll grant you we didn't have to IPL the system every day and we did average 99% uptime during prime shift. On graveyard shift when the systems programmer make their changes it was another story but since the system wasn't officially available during those times we didn't have count the downtime in our stats.
No, there are only 3 differances that I can think of that really make differance
1) A Support contract - We pay big bucks and when there is real trouble truly amazing things will happen. They almost were going to rent a plane just to fly a part to us once because we were down with a hardware failure and we were approaching the 24 hour limit in our support contract.
2) Strict control over hardware and to a lesser extent the software on the mainframe. If some issue of hardware or software arose and fault was serious enough IBM could void the support contract unless it was fixed. In fact a bug was was found once in the OS and IBM contacted us and said that you must install this patch to to the OS or you will no longer be under support.
3) Finally, the complexity and 24/7 nature of computing today has really stressed machines far more than we used to back in the good old days. While we didn't IPL every day we did "roll the regions" every night starting at 4:00 AM just to make sure things would run smoothly the next day. Now we can't do that because we've got a region that needs to be up 24/7 to support the web.
