About the Open Hub Server Farm
Open Hub has a farm of servers that continually monitors source control systems for changes, and creates updated project reports as those changes are discovered.
Given the large number of projects Open Hub is monitoring, we are unable to make real-time updates to project reports. Our goal is to update every project every day, but we are typically trailing by several days.
There is no fixed queue that determines when each project will be updated. For the curious, a detailed explanation of update scheduling can be found below.
Although we can’t publish a precise schedule, we can give you a general idea of how quickly the update queue is currently running. The following graph (updated hourly) shows the creation date of the current report for every project.
Project Reports Updated By Date
How Updates are Scheduled
We have a large farm of servers, and the source code for an individual project is stored redundantly on a random selection of 2 or 3 servers. For efficiency reasons, project updates only run on machines that already have a copy of the project source code. When an individual server becomes idle, it checks its collection of source code for the project with the oldest report, and updates that project.
Because projects are distributed across our servers, and because we can’t predict which server will become idle next, we can’t easily predict exactly when a particular project will be updated.
New projects have priority over existing projects. If any new projects exist, a server will attempt to download a new project instead of updating one of its existing projects. If a lot of large new projects are added to the system, it might be a long time before a server has a chance to update an older project.