Hi,
We plan to make an update from OJS 2.4.8.1 to OJS 3.1 but how will it perform? At the time, when we’ll make the update our platform features over one hundred journals. Kevin Stranach wrote earlier:
”OJS 3.0 will perform best for open-access publications with small to medium-sized amounts of content. As the new codebase matures with follow-up releases, we will ensure its performance with larger collections and higher traffic volumes.”
Therefore I’m a little bit worried. Should we wait for a later edition?
Regards
Niels Erik
Hi @nef,
We’re aware of a few sites working with larger data sets – @ajnyga, for example, is involved with http://journal.fi/, and @alexxxmendonca is working to migrate Scielo. We aren’t currently doing performance testing with large sets internally, but we’re interested in working with groups to identify bottlenecks if any appear. OJS 2.x and 3.x share a lot of the same underlying technology and will have similar performance characteristics, with a few exceptions – such as OJS 3.x’s increased use of AJAX – so we’re not expecting many surprises.
Regards,
Alec Smecher
Public Knowledge Project Team
Thanks Alec, but then why did Kevin write the above?
Regards
Niels Erik
Hi,
Just a brief comment to @nef that we have indeed used 3.0 since January on a installation with aroun 50 journals. Some of the journals date back to 2006 and OJS2 so there is a fair amount of content in Journal.fi. Hard to say how big it is in context of other similar services.
I have not noticed any unusual performance issues in the frontend. In the backend there is a clear difference between OJS2 and OJS3 due to the point that Alec made about OJS3 using more Ajax. But I think that this is something that does not change if you compare a small installation to a larger one. Our production server is just as quick as my own test server with zero traffic.
Hi @nef,
I actually wrote the text you quoted above along with Kevin. We suggested Open Access publications because OJS 3.x won’t include payment/subscription support until 3.1 is released, and wanted to add a caveat regarding performance because of the lack of higher volume testing on the initial 3.0 release. We depend on community feedback for a number of things, often including performance, and now that numerous sites are using OJS 3.x we’re able to have a better sense of production performance and bottlenecks. And as always, if you do encounter a performance bottleneck, we’re willing to work with you to identify and resolve it.
Regards,
Alec Smecher
Public Knowledge Project Team