Doubt about performance degradation (overhead) using EVMS

Hi to all:

I have tested several EVMS 2.x volumes with bonnie, iozone and dd tools to
obtain thoughput and timemarks. I have observed that times are almost
identical when using directly a plain device (ex. /dev/sda) or use this
device contained in a EVMS volume (/dev/evms/evms-01).

I have test several sizes for the volumes (100GB, 1TB, 4TB) with ext3, and
with 0% size used, 50% size used and 90% size used, and every time tests
were very similar using the plain device and the EVMS volume.

Which is the treatment for metadata and indirections to obtain this results?
Are the indirect access cached in memory?, what is the real overhead that i
can suffer using EVMS? Could the performance be degradated when volumes grow
with resizing or having a high use?

Thanks in advance.

Andrés.



----------------------------------------------------------------------------
-----
Andrés Marchante Tirado.
Responsable de Unidad de SVA y Sistemas.
Area de Ingeniería y Tecnología.
-----------------------------------------------------------------
Sociedad Andaluza para el Desarrollo
de la Sociedad de la Información S.A.U. 
Consejería de Innovación Ciencia y Empresa
JUNTA DE ANDALUCÍA
Avda. de la Arboleda s/n
41940 - Tomares, (Sevilla)
Teléfono: 955 06 39 33
Móvil: 697 955 536
e-mail: [email protected]

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

_______________________________________________
Evms-devel mailing list
[email protected]
To subscribe/unsubscribe, please visit:
https://lists.sourceforge.net/lists/listinfo/evms-devel