I received the following email via our press email address. It seems
slow election processing in Estonia was blamed on the Postgres database.
Any idea how to respond to this?
---------------------------------------------------------------------------
Hello,
There was parliament elections in Estonia last sunday. Everything else
went fine except the system lagged for most important hours during
entering the votes from the departments.
Helmes (www.helmes.ee), the developer of many other national systems,
released a note to media on Monday accusing Postgresql database that it
did not manage with the dataflow. Although to calculate for a second, then
about 600 inserting people from the election departments who pass totally
50'000 rows to tables during two hours is a very slow day for most
sql-systems. They mainly need only the next information for voting -
department, candidate, total votes. So the real fault was in the brain of
the database developer - it is silly to say subaru is bad if you cant
drive rally-style and crash the car in the first corner.
I hope Postgresql management will take action on this and give a reply in
media, as accuses were against "freeware databases" - issues like this
give them bad reputation and these databases wont be taken seriously the
next time. I have used postgre and mysql for ten years to build up
photobanks with millions of pictures inside and developed other large
systems and met no mistakes, it all begins with good database design.
News headlines in estonian concerning this topic:
http://www.delfi.ee/news/rk/uudised/helmes-viga-oli-vabavaralises-andmebaasimootoris.d?id=41628665
--
Bruce Momjian <bruce@momjian.us> http://momjian.us
EnterpriseDB http://enterprisedb.com
+ It's impossible for everything to be true. +
