Consider `REP 1 REP 1` placement (selects/filters are omitted).
The placement is `[1, 2], [1, 0]`. We are the 0-th node.
Node 1 is under maintenance, so we do not replicate object
on the node 2. In the second replication group node 1 is under maintenance,
but current caching logic considers it as "replica holder" and removes
local copy. Voilà, we have DL if the object is missing from the node 1.
Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
The node can have MAINTENANCE status in the network map, but can also be
ONLINE while responding with MAINTENANCE. These are 2 different code
paths, let's test them separately.
Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
We already provide the pool and this argument is used only for
preallocation. No functional changes.
Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
* Introduce common method EverExisted
* Define DeletionInfo for struct that must implement Source
* Refactor tree srv
Signed-off-by: Airat Arifullin <a.arifullin@yadro.com>
Current implementation has some quirks. For example,
using only half of object.put.pool_size_remote threads
tells replicator that is node is 50% loaded,
but in reality we could be putting lot's of big objects.
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
Previously, we can continue to return `EndOfListing` infinitely.
Reflect iterator reuse via Rewind() method.
Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
