I almost never experience a register crash. Usually the worst we get is a register locks up for several minutes.
We DO have issues with the Credit Card network. Several times a month something goes wrong with the connection. Most common is that it will stop taking Debit, but sometimes it stops taking all cards. This will last for hours, but never for long enough that a tech actually arrives.
Recently, techs came and replaced our routers, and the problem has been much better since then. He said the old equipment wasn’t compatible with our software.
The store in question is part of a national chain with proprietary POS software that is made and run by a major vendor. We have 2 registers per store, and our store averages between 2 and 3 items per transaction.
At my last store (same company), we did have a recurring crash problem.
One of the registers would randomly drop off the network. The problem was that the other register gave no indication this had happened, so if you were only using one register you wouldn’t know.
When it lost the network, it couldn’t accept cards or checks or access the safe, and the only way to get it to reconnect to the network was to reboot it.
If you ran end-of-day while one register was off the network, it would shortly thereafter cause the register that was on the network to produce an error message saying that had happened, and then it would refuse to do anything until you entered a supervisor ID # from the Help Desk.
So one register would lock up while the other was being rebooted.
I keep that supervisor ID # in a safe place so that I always have it, because there is nothing quite like looking at a growing line of people you can’t ring up.
Slashdot recently had a link to an article about exploits that might be usable against some of those keyboard wedge barcode scanners. Execute commands via a barcode, fun for the whole family.
Back to the story related in the OP. On Black Friday every bit of the entire chain is under a max case load test. Everything from the individual registers to the store level back-end servers to HQ’s servers, plus all the interconnects to the credit and debit processors and to check approval/acceptance companies are all under great load.
Plus all the e-commerce which still flows to the same HQ level servers and all the same credit and debit processors.
Cloud system vendors are all about the idea that you can turn up extra servers easily for little incremental cost and zero incremental capital investment. Which is great until every single customer of the cloud vendor wants to do that for the same few days.
Bottom line: Stuff failing on Black Friday is more about unusually heavy loads than it is about inherent daily bugs.
