From: Judy Miriga

Folks,

This is the most fake explanation I have heard in my life time.

Judy Miriga
Diaspora Spokesperson
Executive Director
Confederation Council Foundation for Africa Inc.,
USA
http://socioeconomicforum50.blogspot.com

— – – – – – – – – –

— On Wed, 3/6/13, roz kahumbu wrote

IEBC Tech Kenya

4am March 6 update from IEBC
http://iebctechkenya.tumblr.com/post/44698659119/4am-march-6-update-from-iebc

Following up on the earlier IEBC press conference (4AM), here are the technical details of the problems:

– Disc space on the server. While we all tend to roll our eyes, ‘how can they run out of disk space’, the reality is that they ran out of configured disk space. In the rush to set up everything before March 4, the IT staff did not configure the server partitioning correctly. Its like having two mailboxes next to each other to receive data but when one fills up it does not correctly overflow to the other one. That error has been fixed.

– Data is stored on the main server at Bomas (which was delayed), a backup server off site (which did not have the new data because it was not on the main server), and then servers at the constituency and county levels. So data probably exists outside Nairobi but it has to be pulled, with scripts, to Nairobi. They are currently checking to see how much exists that is not on the server in Nairobi. It is also not advisable to pull it all at the same time because then the reporting percent will jump from 28% (at 0725) to a much higher number. So it has to be done incrementally.

– The second problem is the transmission of data. KTI’s Deputy Country Representative was at the Farasi School polling center in Westlands Constituency at 2100 on the 4th. He observed the counting of the presidential results and then the box was sealed and the Officer tried to submit the results. He was watching the screen and it was unable to send because of a data problem. Unknown to him, this is the same time that the server went down. The clerk was supposed to only try twice and then move to counting the next box. Instead he kept trying. He could not even enter the results for future transmission because he had not used the phone the night before to connect to the server and to download the data of who was running at that polling station. Also note that the officers and clerks at that station had been on their feet since 5AM and by 9PM had only counted one box of six.

– The third problem is the network coverage. If there is no network coverage, the counting of all 6 boxes must be completed before the officers can move to the constituency office and transmit the data en-route or whenever they have coverage. In the design, the data is supposed to be entered at the polling station even if it cannot be transmitted; then the data is queued for transmission when there is a signal. Based on the example above (Farasi Primary), the data could not have been entered because it was not correctly set up at the beginning of the day.

The IEBC call center is calling all constituencies asking them to re-transmit the data. Data can be re-transmitted many times since it is keyed to the phone and polling center. It cannot be changed once transmitted the first time but multiple transmissions are accepted.

Observation

The returns were 28.21% at 7:25AM, at 7:50 (25min) it is 28.94%. This represents 243 polling stations (of the 33,400 stations). The system is working, Kenyans need to be patient. As one person from the IEBC stated, ““Its too early to be celebrating or commiserating”

March 6, 2013 (5:12 am)
#kenya
#iebc
#polling
#election
#media

—
Jobs in Africa – www.wejobs.blogspot.com
International Jobs – www.jobsunited.blogspot.com