Hello everyone,
I’m considering implementing Open Journal Systems (OJS) in a high availability environment and am wondering if anyone has experience with this. Is it feasible to replicate the OJS database across multiple servers and store documents on a network in a shared server, using a load balancer to distribute the loads? I would greatly appreciate any advice or shared experiences on this topic.
Thank you,
Hi @sacandia
The only fellow I know are doing something like this are @AndrewGearhart.
Andrew, no hurry but I’m sure the community will appreciate you sharing two lines sharing your experience.
Cheers,
m.
I’d like to say that we’re in a high-availability mode, however, we are still battling with a few demons in our system. We have deployed our database in a clustered mysql database arrangement, however, there is not a HA configuration setup for our OJS web instances. Things that would need to be addressed include verifying that session/state management would be handled in the database and not via something specific to the web server. Files are another concern. Currently, files are placed on the file system directly by OJS. This would mean that the file system would need to be handled in a distributed manner. With other content management systems, we’ve run into issues whereby the temp storage during file upload must be shared in addition to the actual file storage. This makes things… complicated. If OJS could be generalized to have file uploads handled with a post to an s3 object store, things could get more successfully segment and be more available for HA. That’s not a current feature that I’m aware of in OJS though. So, to succinctly answer your question - “ish”. 
Hope this delayed reply finds you well. -A