What did happen was that as the crew of about 20 memory card uploaders, each standing at a Tsx machine networked into the server/tabulator, started uploading one at a time, and a precinct at a time, beginning the uploading of 5,793 memory cards from the 1436 precincts, soon the server crashed, bringing down all the Tsx devices too.
L- A look into the tabulation room, through the glass, to lines of Tsx devices with workers uploading memory cards. Board member, Inajo Chappell on right.
R- after the server crash, the second time, the TSx's went down again too. Blank screens. Workers waiting.
After they backed up the system and the votes, made sure nothing had been lost, they started bringing all the Tsx touchscreen devices back up for more uploading. Soon the system crashed again.
It crashed a total of 3 times. Robert Chen, national manager of the Diebold, DIMS unit (Diebold Information Management, their registration system subsidiary) was there from Sacramento; as well as the Diebold tech working with the CCBOE, Archie ______?
(Pictured above: Robert Chen, middle, the Diebold Edward Howard PR firm rep, Wayne Hill, former communications manager for former OH Sos, Celebreze, Jr. to the right.)Together with CCBOE ballot department and I.T. they tried to figure out the problem.
As Robert Chen explained to me, the problem was one of information "collisions"/ "switching." When I questioned if it was one of "not having enough lanes on a highway," he explained that "no, it was more like not having good enough traffic cops to direct the traffic."
Chen did not seem upset or rattled. In fact during the last two hours, he, Hill and regional sales rep, Barry Herron (sp?) stayed mostly in an office behind the glassed tabulation room, possibly to get out of the way, possibly to rest, possibly to get out of the way of media cameras by then peering into the room, eager to get the news and results stories.
The CCBOE will continue with investigation of what really happened to their multi-million dollar system.
I also overheard Chen questioning if the CCBOE had stress-tested the system, which they said, of course, they had - in every election and often in between.
A good friend and computer expert was amazed to hear this. He said the question to ask Diebold is if THEY had stress tested their system, under a Cuyahoga County load, as they should have before they sold it as fit for Cuyahoga.
He was further amazed that such a basic problem as switching - having a really quality switch on our system - was not taken care of by Diebold long ago. That too, though adding expense to the seller, is basic to such a system, as laid out by Diebold/Premier. Their techs have been there consistently and regularly, helped plan the layout, and the prep.
The solution they jointly came up with after the crashes, was that they lessened the number of Tsx's used, but more, they had workers stop uploading every half hour, and after the last one finished the upload he/she had started, they backed up the votes onto another server (the one borrowed earlier this year from the Secretary of State, because our original server had crashed, and Diebold had them send it around the country, during which it physically crashed (was dropped,) and Diebold had not fixed it) they waited a few minutes, and began again. After a couple of these rounds of stopping and backing up, they finished at about 1:30 am.
But...then some larger problems came to light....
It was then that they found the few precincts with missing memory cards - which were found within 15 minutes, because the supply "transfer cases" had been put away in an orderly fashion after being brought in. Those poll workers had stored their memory cards for return in the wrong bag. CCBOE workers knew exactly where to go find the missing cards.
throughout the then-halting upload process, it was found that 8 memory cards had come in blank - were somehow corrupted.
The reasons for that, according my computer guru friend, could be many, and could be very worrisome to election security. What would corrupt those memory cards? and why? and why was there no warning to poll workers?
This is very similar to the same problems reported by Election Science Institute in August of last year, when as hired by the Commissioners, they audited the debacle May election, and found corrupted cards, that gave no warning - a fact which was brushed aside, then later, as if waving a magic wand, was tried to be declared "untrue" by Mr. Bennett, Vu, Diebold, and the previous board.
Not "insurmountable" "glitches"?
As the memory cards were found blank, and to get a vote count, Platten began sending out crews to polling locations of each of those cards, to pick up all the Tsx devices from those locations. The CCBOE was going to have to set up all the devices, per location,(up to 15 machines each locale) to find the correct machine for the corrupted card, to first of all get the votes from that machine's archived memory.
When I left at about 1:45am they were just emptying the take-in area to begin that machine set-up/the matching process to get the rest of the votes.
Also when I left they had only reached 4 of those locations to regain access that night to pick up those machines.
At 1:45 am, and with worker exhaustion setting in, it was being found impossible to get the machines from the other 4 locations, schools and churches, until a more reasonable time that morning. They were forced to be there all night, and not finish until 11:30 am Wednesday - for a wonderfully run election, that by all rights should have completed at 11:30 pm. Tuesday night, all because of Diebold failures.
In Thursday's PD, a Diebold/Premier spokesman, Chris Riggall, (is it any wonder they could not sell their company, and had to spin it off under another name?) suggested votes were being downloaded "too fast."
Aside from the fact that votes were being uploaded, not downloaded, it's hard to even reply to such self set-up, computer company ridiculousness. He also said "It's not OK, and it's not insurmountable."
No skin off his nose. And then why didn't Diebold "surmount" long before?
He also said, "It's not how you want your vote tabulations to go." No, Mr. Riggall, these are your tabulations, and you guys screwed up ...still and again.
The PD also reported that SoS Brunner said election workers did not cause the glitches. (I really hate that word. These are not glitches. They are major, ongoing, recurring serious problems to our election results, with manufacturers taking NO responsibility, while no one with the authority to do so, like the Secretary of State, holds them accountable to the boards and to taxpayers who've paid them multi-millions for such debacles.)
The only good that may come out of this very public Diebold operational failure, is that it may finally prove to be the last straw against Diebold. After years of notices of their money-hungry actions, coupled mostly with little reliability, and the numbers of security holes also built into their election systems, it's time to rid ourselves of this system and this vendor.
With Brunner's statewide testing due to be done mid next month, it's certainly got to be time.
This country cannot afford another Florida 2000, and Ohio 2004, in 2008 - with people like Diebold driving the handbasket.