Archiva performance question

[Sergey Nikitin]

Hi,

I've started with Archiva evaluation and noticed that build time for project is increased by ~40-50% in case Archiva is used instead of simple remote repositories specification.

I am building my test project on "machine A", Archiva is installed on "machine B". Both machines are in a very fast network.

Typical maven behaviour is:

"Downloading: ..."
hang for 2-3 seconds
"... downloaded"

File size don't matter that much - POM file and 6Mb jar got almost same download time from Archiva.
Internet connection is very fast for both machines.

Are there some things to take a look on? Or it is inevitable performance drop due to all the things Archiva does behind the curtain?

Sergey

[Julien HENRY]

Hi,

In my company some users are reporting the same issue: they prefer to use direct repository access instead of our internal Archiva because of performance drop.

I plan to use an external DB instead of embedded Derby to see if it improves performances, but I'm very interested to know if there is something else I can do.

Regards,

Julien




________________________________
De : Sergey Nikitin <nikitin.k172@...>
À : users@...
Envoyé le : Mardi, 21 Juillet 2009, 17h55mn 49s
Objet : Archiva performance question


Hi,

I've started with Archiva evaluation and noticed that build time for project
is increased by ~40-50% in case Archiva is used instead of simple remote
repositories specification.

I am building my test project on "machine A", Archiva is installed on
"machine B". Both machines are in a very fast network.

Typical maven behaviour is:

"Downloading: ..."
hang for 2-3 seconds
"... downloaded"

File size don't matter that much - POM file and 6Mb jar got almost same
download time from Archiva.
Internet connection is very fast for both machines.

Are there some things to take a look on? Or it is inevitable performance
drop due to all the things Archiva does behind the curtain?

Sergey
--
View this message in context: http://www.nabble.com/Archiva-performance-question-tp24590470p24590470.html
Sent from the archiva-users mailing list archive at Nabble.com.

[Sergey Nikitin]

Hello,

Yesterday i've tried Archiva with Oracle and Hsqldb as artefacts database.

Oracle was not working for me due to problems with varchar2 size (Archiva tries to make a column with varchar2(4000+), which is not possible). Few workarounds haven't worked, so i just left it without digging deep.

Hsqldb is ok, but performance win was about 5%.

Im still looking forward to any information concerning my initial question

Sergey

Julien HENRY wrote:

Hi,

In my company some users are reporting the same issue: they prefer to use direct repository access instead of our internal Archiva because of performance drop.

I plan to use an external DB instead of embedded Derby to see if it improves performances, but I'm very interested to know if there is something else I can do.

Regards,

Julien