Currently, under a mixed load one failed PUT can lead to closing
connection for all concurrent GETs. For PUT it does no harm: we have
many other nodes to choose from. For GET we are limited by `REP N`
factor, so in case of failover we can close the connection with the only
node posessing an object, which leads to failing the whole operation.
Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
Currently, DELETE service sets tombstone expiration epoch to
`current epoch + 5`. This works less than ideal in private networks
where an epoch can be e.g. 10 minutes. In this case, after a node is
unavailable for more than 1 hour, already deleted objects have a chance
to reappear.
After this commit tombstone lifetime can be configured.
Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
If we had lots of domains in one zone, `dump-hashes` for all others
can miss some domains, because we need to restrict ourselves with _some_
number.
In this commit we use neo-go sessions by default, with a proper
failback to in-script iterator unwrapping.
Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
We rarely need to list all containers: as one example
we need it for tree service synchronization once per epoch.
Given that cache TTL has the order of block time it makes no sense
to cache the list of all containers.
Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
In NeoFS containers can be removed on behalf of its owner only. To
improve user experience, there is a need to add ownership check to the
removal command of the NeoFS CLI.
Check container ownership in `container delete` command `Run` function.
The check can be skipped by `--force` option.
Signed-off-by: Leonard Lyubich <ctulhurider@gmail.com>
Could be related to "websocket users limit reached" on the `neo-go` server
side when an SN/IR is rebooting repeatedly.
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
If an external session is provided and is not opened by CLI itself, add
children objects to it too. It fixes "not found" errors when removing a big
object with a predefined session (from a file).
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
If a `neofs-cli object delete` operation is performing using a bearer token,
add it to the new `HEAD` requests that collects children OIDs.
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
Do not use node's local storage if it is clear that an object will be
removed anyway as a redundant. It requires moving the changing local storage
logic from the validation step to the local target implementation.
It allows performing any relations checks (e.g. object locking) only if a
node is considered as a valid container member and is expected to store
(stored previously) all the helper objects (e.g. `LOCK`, `TOMBSTONE`, etc).
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
In previous implementation node lost maintenance status after successful
switching to it. For example, after some period of time node sent
bootstrap requests with the "online" state instead of "maintenance".
Make `startMaintenance` method to set maintenance status in the
`networkState`.
Signed-off-by: Leonard Lyubich <ctulhurider@gmail.com>
In previous implementation bootstrapping state was chosen according to
bool flag which was not convenient.
Create separate method to boostrap with "online" and the current state.
Signed-off-by: Leonard Lyubich <ctulhurider@gmail.com>
Object removal session should reflect all objects related to the
removing one.
Make `OpenSessionViaClient` to gather the split members of the original
object in order to spread the session to them.
Signed-off-by: Leonard Lyubich <ctulhurider@gmail.com>
Add new shard modes as a map entry to automatically parse them in
`set-mode` command. The change also automatically adds new modes to help
message.
Signed-off-by: Leonard Lyubich <ctulhurider@gmail.com>
There is a need to support NeoFS-binary sessions along with JSON ones in
NeoFS CLI.
Provide generic `common.ReadBinaryOrJSON` functions which tries to
decode NeoFS-binary structure and falls back to JSON format. Use this
function in all places with token reading.
Signed-off-by: Leonard Lyubich <ctulhurider@gmail.com>
In previous implementation turning to maintenance mode using NeoFS CLI
required NeoFS API endpoint. This was not convenient from the user
perspective. It's worth to move networks settings' check to the server
side.
Add `force_maintenance` field to `SetNetmapStatusRequest.Body` message
of Control API. Add `force` flag to `neofs-cli control set-status`
command which sets corresponding field in the requests body if status is
`maintenance`. Force flag is ignored for any other status.
Signed-off-by: Leonard Lyubich <ctulhurider@gmail.com>
Node's maintenance state can be set only if the network allows it.
Make `neofs-cli control set-status` to read current network settings and
check is maintenance state is allowed. Fail the execution if the mode is
not allowed.
Signed-off-by: Leonard Lyubich <ctulhurider@gmail.com>
Add `SynchronizeAllTrees` method of the Tree service. It allows fetching
tree IDs and sync all of them. Share common logic b/w the new method and
the `SynchronizeTree`.
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
It allows general usage of `controlNetmapStatus` getter: does not matter of
handler (`NewEpoch` handler of initial setter).
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
In previous implementation node always sent bootstrap requests with
"online" network state. This provoked switching to online of the nodes
under maintenance.
Signed-off-by: Leonard Lyubich <ctulhurider@gmail.com>
Storage node should not provide NeoFS Object API service when it is
under maintenance.
Declare `Common` service that unifies behavior of all object operations.
The implementation pre-checks if node is under maintenance and returns
`apistatus.NodeUnderMaintenance` if so. Use `Common` service as a first
logical processor in object service pipeline.
Signed-off-by: Leonard Lyubich <ctulhurider@gmail.com>
Make it store its internal `zap.Logger`'s level. Also, make all the
components to accept internal `logger.Logger` instead of `zap.Logger`; it
will simplify future refactor.
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
It does not make sense to open remote sessions with the storage node in
`get`, `head`, `search`, `range` and `hash` sub-commands of `neofs-cli
object` command.
Do not use NeoFS API `SessionService` in mentioned commands. Decode
object session from JSON file specified `--session` flag. Perform some
sanity checks instantly on CLI side.
Signed-off-by: Leonard Lyubich <ctulhurider@gmail.com>
If the contract returns a netmap that does not contain the node, update
local `NodeInfo`. It fixes `neofs-cli netmap nodeinfo` command that printed
"state: online" previously.
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
Currently, when removing shard special care must be taken with respect
to shard numbering. `mode: disabled` allows to leave shard configuration
in place while also ignoring it during initialization. This makes
disk replacement much more convenient.
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
In previous implementation of `neofs-node` app object session was not
checked for substitution of the object related to it. Also, for access
checks, the session object was substituted instead of the one from the
request. This, on the one hand, made it possible to inherit the session
from the parent object for authorization for certain actions. On the
other hand, it covered the mentioned object substitution, which is a
critical vulnerability.
Next changes are applied to processing of all Object service requests:
- check if object session relates to the requested object
- use requested object in access checks.
Disclosed problem of object context inheritance will be solved within
Signed-off-by: Leonard Lyubich <ctulhurider@gmail.com>
Split all the fields in `cfg` structure on:
1. `applicationConfiguration`;
2. `internals`; // shared entities for an application work, such as
`context.Context`
3. `shared`; // holder for the shared entities b/w;
4. `cfgXXX`; // configuration for internal services.
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
In previous implementation node blocked any operation of local object
storage in maintenance mode. There is a need to perform some storage
operations like data evacuation or restoration.
Do not call block storage engine in maintenance mode. Make all Object
service operations to return `apistatus.NodeUnderMaintenance` error from
each local op.
Signed-off-by: Leonard Lyubich <ctulhurider@gmail.com>
Make `netmap snapshot` command to print `MAINTENANCE` state of the nodes
with `IsMaintenance()` flag set.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
After recent Netmap contract changes all read methods which return
network map (either candidates or snapshots) encode node descriptors
into same structure.
Decode `netmap.Node` contract-side structure from the call results.
Replace node state with the value from the `netmap.Node.State` field.
Signed-off-by: Leonard Lyubich <ctulhurider@gmail.com>
Make storage node to return `NODE_UNDER_MAINTENANCE` status
error on each local object operation if the node is in `MAINTENANCE`
mode.
Pass `apistatus.NodeUnderMaintenance` to `StorageEngine.BlockExecution`
during `ControlService.SetNetmapStatus` RPC processing.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
After recent changes `MAINTENANCE` state is reflected in the Sidechain.
Storage node should switch its state to "maintenance" during serving the
`ControlService.SetNetmapStatus` RPC with correspoding status in the
request.
Call `UpdatePeerState` operation of Netmap contract's client in
`control.NodeState` provider on Storage node app side. The op is
executed if `BlockExecution` on local object storage is succeeded.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
After recent changes network configuration provided by NeoFS storage
nodes contains `MaintenanceModeAllowed` flag. There is
a need to support this value in NeoFS CLI application.
Print `MaintenanceModeAllowed` flag in `netmap netinfo` command.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
After recent changes network configuration stored in the Netmap contract
of the NeoFS Sidechain contains `MaintenanceModeAllowed` flag. There is
a need to support this value in Storage node application.
Make `NetmapService.NetworkInfo` RPC server of the storage node to set
`MaintenanceModeAllowed` flag according to corresponding value in the
Sidechain.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
`NetworkConfiguration` represents NeoFS network configuration stored in
the Sidechain. In previous implementation the configuration missed flag
of disabled homomorphic hashing.
Add `NetworkConfiguration.HomomorphicHashingDisabled` boolean field.
Decode the field in `Client.ReadNetworkConfiguration` method. Print this
value in `netmap netinfo` command of NeoFS CLI.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
In previous implementation storage node interpreted all status values
sent in `SetNetmapStatus` RPC as `OFFLINE` except `ONLINE` and
`MAINTENANCE`. This could lead to incorrect processing of new values,
and also didn't allow detection of problems with sending garbage values.
Make implementation of `NodeState` interface used by Control API server
to deny requests with statuses other than protocol-declared enum.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
It will allow rereading config values and will simplify distinguishing them
from the custom values in the `cfg` structure.
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
Negative values have no sense. On the other hand it differs from the
blobovnicza's configuration and prevents unification.
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
Replace `ProcessCurrentNetMap` method of `NodeState` interface with
`ReadCurrentNetMap` one with two changes:
* Replace network map type from NeoFS SDK package with the
protocol-generated message. This replaces all the business logic to
the application layer.
* Support error return. This allows to cover problem node states.
Return an error from `NodeState.ReadCurrentNetMap` method implemeted
through `atomic.Value` if `Store` method has not been called yet.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
1. Do not require a request to be signed by the container owner if a
bearer token is missing
2. Do not check the system role since public requests are not expected to
be signed by IR or a container node (unlike the object requests)
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
There is a need to have the ability to track NeoFS timeline on storage
nodes. Epochs tick on notifications receipt, so the most obvious way to
know about received epochs is logging the events.
Wrap `morphEvent.ParseNewEpoch` event parser into function which writes
log message about new epoch number.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Implement methods that `actor.Actor` requires:
1. `InvokeFunction` -- wrapper over `InvokeScript`
2. `GetVersion` returns default struct
3. `CalculateNetworkFee` copied and simplified from Neo-go server side.
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
NNS proposal describes string N3 address format, however we must also
have hex-string for backwards compatibility.
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
In case we already have the record and it is invalid, we should
overwrite it instead of having several conflicting records.
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
NEO NNS proposal uses addresses. We should eventually use the same,
but must stay compatible now.
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
Flag `--pre-check` of `set-eacl` command found to be in demand in most
cases. based on this, it makes sense to add its action to the default
behavior.
Pre-check container extensibility by default. Rename flag to
`--no-precheck` and invert its action.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Container ACL in NeoFS can be extended only for container in which the
corresponding option is enabled. In previous implementation command
`set-eacl` could hang up on modifying eACL of the non-existent or
non-extendable container. To improve UX, there is a need to pre-check
the availability of `SETEACL` operation.
Add boolean `precheck` flag to `set-eacl` cmd which reads the container
before the actual transaction formation. If flag is set, command fails
on non-extendable container ACL.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
If container listing cache on node's side is missing (for particular
owner), then updating it as a reaction to successful container creation
leads to potentially invalid cache value for a period of time equivalent
to cache TTL.
Immediately return from `ttlContainerLister.update` method if owner's
container list isn't cached.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
From now cache TTL can be parameterized in the `neofs-node` app using
`cache_ttl` config key. `disable_cache` value is no longer supported.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
In previous implementation storage node responded with the outdated
container list after successful creation/removal up until cache
invalidation due to TTL. In order to decrease the probability of
outdated responses node should update its cache on event receipts.
Implement `ttlContainerLister.update` method which actualizes cached
list of the owner's containers. Make node to call `update` method
on `PutSuccess`/`DeleteSuccess` notifications from the `Container`
contract.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
In previous implementation storage node responded with the removed
container up until cache invalidation due to TTL. In order to avoid
false-positive responses node should update its cache on `DeleteSuccess`
events.
Make node to call `handleRemoval` method of the container cache which
leads to subsequent `apistatus.ErrContainerNotFound` errors.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
In previous implementation failed requests to the Sidechain weren't
cached. It makes sense to cache errors along with the values in order to
decrease potential load spikes onto Sidechain nodes.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
There is a need to sync container-related caching mechanism with the
actual Sidechain changes. To do this, node should be able to listen
incoming notifications about container ops.
Define `PutSuccess` / `DeleteSuccess` notification event's parsers.
Subscribe to these events in node app. As initial implementation node
will log event receipts. Later handling is going to be practically
complicated.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
1. Move compression parameters to the `shard` section.
2. Allow to use multiple sub-storage components in the blobstor.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
Define `--with-attr` flag of `container list` which makes the command to
request and print user attributes for each found element.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Define `--with-attr` flag of `container list-objects` which makes the
command to request and print user attributes for each object from the
container.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
We have the default value which is also printed in the help messages but any
call that does not specify that flag leads to an error.
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
`Degraded` mode can be set by the administrator if needed.
Modifying operations in this mode can lead node into an inconsistent state
because metabase checks such as lock checking are not performed.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
Do not check that a node indeed belongs to the container, because the
synchronization will fail in this case anyway.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
Do not return backend type from the service for now, because memory
backend is expected to vanish.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
There is an error only if the flag is not defined, such errors should be
caught during debugging.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
