I was wondering how the admins of TeamCity instances out there handle their storage. By default, TeamCity will keep all artifacts for ever, but that's typically not sustainable.
Reading the documentation about Clean-Up (http://confluence.jetbrains.com/display/TCD8/Clean-Up), a reasonable configuration seems to set a time or build-number based cleanup policy, together with pinned builds, to keep important builds, like the released ones.
Still, over time the number of pinned builds will grow, and the storage requirements along. If the artifact storage is stored on a simple local partition, it won't be easily scalable. Is it recommended or supported to put the artifacts on a network share for better scalability? I wonder how it impacts the overall performance.