Re: [OT] Any major users of postgresql?

Поиск
Список
Период
Сортировка
От John Moore
Тема Re: [OT] Any major users of postgresql?
Дата
Msg-id DT017.3226$G_1.332362@newsread2.prod.itd.earthlink.net
обсуждение исходный текст
Ответ на [OT] Any major users of postgresql?  (Joe Brenner <doom@kzsu.stanford.edu>)
Ответы Re: Re: [OT] Any major users of postgresql?  (Tom Lane <tgl@sss.pgh.pa.us>)
Список pgsql-hackers
> It probably depends on what you call "serious". Anyway, the project I am
> working on is a online community for alternate investments and is built
> around a PostgreSQL (first 7.0, now 7.1) database: it's
> <http://village.albourne.com> but unfortunately most of it is limited
> only to subscribers so there is not a lot db-related to see. It's
> PostreSQL + Apache + mod_perl on Digital Unix.
>
I would define "serious use" as use in transactional applications where the
loss of data input by users is a very bad thing, and the uptime requirements
are 24x7, with availability requirements overall of >99%.
<p>
As an example, and where most of my past experience has been, consider a
reservation system for an airline or hotel chain. Such a system may have
hundreds to thousands of transactions per second. More importantly, tens per
second of those transactions which must not be lost - i.e.
reservations/changes/cancellations - and which are worth real money. Losing,
say, 15 minutes of these is a catastrophe. Note also that transactions are
not equivalent to page views - in this case a single "page view" would
result in a series of many database operations to generate a single
response.
<p>
An even tougher example would be an online financial system such as an ATM
debit system. In that case, you can hand someone a lot of money as a result
of a transaction. Loss of that data is exactly loss of the money!
<p>
In the case of PostgreSQL, as far as I can tell, one could lose all data
since the previous dump if one lost the database media. In Oracle or
Informix, that is *not* true, because they can do a point-in-time restore
from the last full save, based on the WAL's.





В списке pgsql-hackers по дате отправления:

Предыдущее
От: Rachit Siamwalla
Дата:
Сообщение: varchar vs. text
Следующее
От: Bruce Momjian
Дата:
Сообщение: Re: SOMAXCONN (was Re: Solaris source code)