Hello Felix, firstly, thanks for your email we would more than happy to change it, but the opennebula driver is managed by OpenNebula guys, so surely they're looking at this mail and they will be more than happy to fix the issue.
Obviously, if you want, you can send a patch if you think would solve the issue and if match the criteria will be accepted :) Cheers, Francesco Vollero On Mon, Jun 25, 2012 at 5:29 AM, Felix Lee <[email protected]> wrote: > Dear All, > I am new with deltacloud, for integration requirement, we are evaluating > deltacloud solution. > Here, I don't know if anybody ever reported this issue, however, recently, I > noticed deltacloud is in poor performance with our open nebula instance. > Usually, our opennebula always runs over 500 VM instances, and while we use > deltacloud to query instances, it will take pretty long time to get all > results. > > > Here are some examples: > > With "deltacloudc instances" command: > > real 20m19.293s > user 0m7.842s > sys 0m0.681s > > > With curl to deltacloud rest api "/api/instances": > real 4m35.891s > user 0m0.002s > sys 0m0.005s > > With curl to opennebula OCCI interface "/compute" directly via the same > host: > real 0m1.249s > user 0m0.001s > sys 0m0.006s > > > It seems to me that the REXML will be the performance bottleneck in > opennebula_driver, so, would you consider to replace it by another one? > > > > BTW, I am using following deltacloud version: > deltacloud-client (0.5.0) > deltacloud-core (1.0.0) > > > And my ruby version is 1.9.3p194. > OS is Scientific Linux 5.8 with 2.6.18-308.8.2.el5 kernel. > > > Best regards, > Felix Lee ~ > -- > Felix Lee Academia Sinica Grid & Cloud. > Tel: +886-2-27898308 > Office: Room P111, Institute of Physics, 128 Academia Road, Section 2, > Nankang, Taipei 115, Taiwan
