Closed vtardia closed 9 years ago
I'm not able to reproduce this anymore. I think the problem was that the seneca initialisation was inside a before function in the test. As part of moving the tests to lab I've already moved that out (see https://github.com/senecajs/seneca-shard-store/pull/12).
Another possible cause is the presence of seneca-store-test at versione 1.0.0 during your tests. Current implementation of shard-store has another serious problem with that versions (see #11)
In current version, with seneca 0.5.20, code like this makes sense:
Every shard has its own seneca instance that uses its specific store and can be called with
shard.seneca.act(args, cb)
in store-specific methods (load, save, etc).By updating dependencies and using seneca 0.7.1 the same code generates an infinite loop. The seneca instance attached to each shard does not use that shard's specific store, it uses the highest level
shard-store
, so each call toshard.seneca.act(args, cb)
recalls itself recursively.Is there a way to address a specific store with the
act()
method?