Fix a bug in the selection when removed object that matches search query
provoked the return of an empty result.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Fix a bug in the selection when an object could be added to the result after
a mismatch in the previous filter.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
In the previous implementation of the metabase, the unique value of the
header was assigned a bucket, the elements of which were leaves with a
key-address and an empty value. This approach was relatively efficient in
terms of write speed. However, a large number of buckets led to a rapid
increase in the database volume (~4GB for 100K objects with unique
attributes). An approach is presented with storing indexes on the value of
headers in the leaves of the tree, where the keys are the unique values of
the header, and the values are a serialized list of addresses (gob
encoding is temporarily used for serialization).
The new approach gave a good result in saving space (~350MB), however, it
significantly reduced the write speed with an increase in the number of
objects (~ 80x after 100K objects).
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
The previous metabase implementation took an exclusionary approach: filters
narrowed the set of all objects to those that match all filters. An
inclusive approach is presented. In it, when traversing the indexed headers,
the object becomes a candidate for selection. If at least one of the
subsequent filters is not passed, the object ceases to be a candidate. At
the end of the traversal, the remaining candidates are added to the
resulting sample. The borderline case of no filters is handled in a special
way: all stored objects are added to the resulting selection.
Presented inclusive approach showed better performance in most scenarios
(although not all).
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
In previous metabase implementation the absence of an attribute presented in
the search filter did not exclude the object from the result. Change this
behavior to exclude the object from the result.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Replace meta Bucket with meta.DB instance in local storage implementation.
Adopt all dependent components to new local storage.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Process parent objects in Put method. Headers of parent object are stored as
regular leaf objects in metabase from now. Build indexes for ROOT, LEAF and
CHILDFREE properties.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Call CheckFormat function in container.Put handler for conducting initial
checks of the structure that are not performed by the smart contract.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
With neo-project/neo-modules#358 nodes support RPC to
return used magic number. Therefore client doesn't need
that configuration value any more.
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
Neo-go does not use smartcontract.Parameter to return values
anymore, so it's convertes partly removed from neofs-node.
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
There is an optimization to send ApprovePeer tx only in
case if `touch` returns false. It returns false if node
is not in the cache or it was flagged to removal.
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
This event will be produced by new epoch handler. All
cleanups should be synchronous because smart contract can't
store inner ring votes forever. Therefore voting should be done
in similar time interval for all inner ring nodes.
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
Cleanup table is a cache for inner ring node to look for netmap
snapshot. It updates access time of bootstrapped nodes and will
be used to send `updateState` txs to clean netmap from unresponsive
nodes.
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
Implement Delete method on DB structure that adds deleted addresses to
tombstone index. Do not attach addresses from tombstone index to Select
result. Return error from Get method if address is presented in tombstone
index.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Implement bolt-based metabase that is going to be used in local object
storage. Implement Put/Get/Select methods.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Make Policer to call AddTask method of Replicator when an insufficient
number of copies of an object is detected in the container.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Implement Replicator service that performs background work to replicate
local object to remote nodes in the container. Replicator is going to be
used by Policer.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
There is a bit to allow or deny bearer token check for
each object service method. If this bit is not set then
ignore bearer token and use extended ACL table from
sidechain.
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
If bearer token is presented in the request then check
if it is a valid one and then use it to process extended
ACL checks.
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
Malicious user can stole public session key and use
it by sending request from it's own scope. To prevent
this each session token is signed and signature private
key must be corresponded with owner id in token. Therefore
malicious node cannot impersonate request without private
key to sign token.
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
Implement Policer service that performs background work to check compliance
with the placement policy for local objects in the container. In the initial
implementation, the selection of the working queue of objects is
simplified, and there is no transfer of the result to the replicator.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Define ErrNotFound error in headsvc package. Return ErrNotFound from Head
method if the header was not found in the container.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Wrong boolean operation order made matcher return false
on `non-root` search query with non-regular objects. Instead
it should return true for `non-root` query and false for `root`
query.
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
If object service produces new request, the should contain
session token. This is the only way for node to grant access
for a private container.
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
Listen and ListenWithError methods check if subscriber channel
has been closed. If so, ListenWithError passes error message
into provided channel.
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
RPC node closes websocket notification channel if it was terminated
or something wrong happened. Subscriber has to check this condition
and alert about this in upper context by closing it's own channel.
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
Inner ring index will be used to access alphabet contracts.
First seven inner ring nods communicating with one alphabet
contract, depending on their index.
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
In #37 we've decided to remove length check, because smart contract would
fail on casting `nil` value from storage to `[]byte` producing FAULT state.
Apparently it does not fail, so we have to check length explicitly.
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
Classifier fetches public key of the request owner
and owner itself. Extended ACL check should rely on
this public key, because it might be extracted from
session token.
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
With one tombstone for split objects we can't simply
place it in container. We should inform all nodes that
store split objects of removed original object.
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
Owner of the request is stored in session token most of the times.
Put request contains session token in the object body, so we have
to fetch it from there.
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
These buckets can be used to store blobs and metadata.
They will be removed as enhanced blob storage will be
implemented for neofs-node. To setup storage type
use `storage.object.type` and `storage.meta.type` params.
Available options:
- inmemory (default)
- boltdb
- filesystem
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
GRPC has default message limit of 4MiB. Since every transmitted
neofs message has to be signed, then original message should
be split into transfer fit structures before signature service.
This commit introduce transport payload splitter for object
service pipeline. This splitter works with stream response
for methods:
- object.Get
- object.Range
- object.Search
Signed-off-by: Alex Vanin <alexey@nspcc.ru>
In previous implementation InitRelations call in payload size limiter was
called in write chunk method. This provoked clearing the split header in
children starting from the second.
Replace InitRelations call to the 1st child allocating stage.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Refactor query to match object and its parents in a single call. Support
KeyRoot and KeyLeaf filters.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
In previous implementation object.Search services allowed to search only
physically stored objects. This limitation did not allow getting the ID of
the split object.
Extend search execution logic with parent object filtering. Parent objects
that passed filters are now included in the result
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Use key storage in object services in order to sign requests with private
session key within user session.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Linking object carries only identifier of the parent object. Fix Validate
method of FormatValidator to not validate parent object of the linking
object.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Add session token argument to object formatter constructor which is written
to the object. Pass session token from trusted object Put.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
