dCache - a system for storing and retrieving huge amounts of data, distributed among a large number of heterogenous server nodes, under a single virtual filesystem tree with a variety of standard access methods
dCache won't deliver a file if there is staging of this file in progress even if there is cached replica available.
Sometimes, if a pool goes down and user requests a file, whose replica was on that pool, for read, a stage request is properly issued by PoolManager. If pool comes back online all attempts to read the file are blocked until stage request is complete. This makes users very unhappy and wastes resource (tape drive time and extra disk replica).
dCache won't deliver a file if there is staging of this file in progress even if there is cached replica available.
Sometimes, if a pool goes down and user requests a file, whose replica was on that pool, for read, a stage request is properly issued by PoolManager. If pool comes back online all attempts to read the file are blocked until stage request is complete. This makes users very unhappy and wastes resource (tape drive time and extra disk replica).
(I opened this issue yesterday in nfs4j repo by mistake https://github.com/dCache/nfs4j/issues/134)