W dniu 14 kwietnia 2010 15:30 użytkownik Grzegorz Jaśkiewicz
<gryzman@gmail.com> napisał:
> That really sounds like hardware issue. The I/O causes the system to freeze
> basically.
> Happens sometimes on cheaper hardware.
>
Probably You have right because this is HS21 Blade Server. And as You
know blades are cheap and good. Why blades are good - because they are
cheap (quoting IBM salesman). I know this hardware is not made for
databases but for now I do not have any other server. Firmware on this
current server is very old and it should be upgraded and there are
many other things to do. VMWare machines (currently I have ESX 3.5,
vSphere 4 is based od 64bit RedHat5 system and is much faster that 3.5
but migration process is not even planned) has still to low
performance for database solutions (of course in using vmdk, not RAW
device mapping or Virtual WWN solution for accessing LUN-s).
As more I am reading than more I see that the file system is wrong
partitioned. For example - all logs and database files are on the same
volume, and that is not right.
Kevin Grittner also mentioned about write back function on the
controller. LSI controllers for blades has that function as far as I
know. I have to check it if that option is turned on.
As I mentioned - I am not familiar with databases so I have made some
mistakes but I am very happy for the effects how fast now Zabbix
works, and how easy PostgreSQL reclaims space. I think it was a good
decision and maybe I will try to interest some people in my company in
PostgreSQL instate of Oracle XE.
Once more time - many thanks to all :)
--
Greetings
Krzysztof Kardas