`log.With` is suitable during initialization, but in other places it induces
some overhead, even when branches with logging are not taken.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
If we should process address based on some condition, there is no need
to read file content in memory.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
Currently we use `(*bbolt.Bucket).Stats().KeyN` for estimating database
size. However, it iterates over all pages in bucket and thus heavily
depends on the bucket size. This commit replaces initial size estimation
with a single `os.Stat` call.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
Core changes:
* avoid package-colliding variable naming
* avoid using pointers to IDs where unnecessary
* avoid using `idSDK` import alias pattern
* use `EncodeToString` for protocol string calculation and `String` for
printing
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
- Delete objects physically on tombstone's arrival;
- Store information about tombstones in the Graveyard;
- Clear Graveyard every epoch based on the information about TS in the
network.
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
Add offset element to the iterations over deleted objects (both the
Graveyard and the Garbage buckets).
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
It allows storing information about object in both ways at the same time:
1. Metabase should know if an object is covered by a tombstone (that is
not expired yet);
2. It should be possible to physically delete objects covered by a
tombstone immediately (mark with GC) but keep tombstone knowledge.
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
Morph "NewEpoch" event handling was registered in a closure over
`addNewEpochNotificationHandler` func. That may lead to the data race:
if a shard was initialized before the event registration, everything works
as planned, but if registration was made earlier, it was not able to
include GC handlers since a shard has not called `eventChanInit` yet and,
therefore, it has not registered handler yet.
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
Also, remove optimization comments:
1. Having to maintain an execute the same logic for headers as for
objects is quite inefficient, as it increases memory footprint.
2. Unmarshaling object is a cheap operation if data slice is in memory.
3. For unmarshaling header-only, I think we need SDK support.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
`Degraded` mode is set automatically after error counter is over the
threshold. `ReadOnly` mode can still be set by an administrator.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
`Batch` can execute the function multiple times leading to multiple
increases of a size approximation.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
`apistatus` package provides types which implement build-in `error`
interface. Add `error of type` pattern when documenting these errors in
order to clarify how these errors should be handled (e.g. `errors.Is` is
not good).
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Replace `ErrNotFound`/`ErrAlreadyRemoved` error from
`pkg/core/object` package with `ObjectNotFound`/`ObjectAlreadyRemoved`
one from `apistatus` package. These errors are returned by storage
node's server as NeoFS API statuses.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
There is a need to process expired `LOCK` objects similar to `TOMBSTONE`
ones: we collect them on `Shard`, notify all other shards about
expiration so they could unlock the objects, and only after that mark
lockers as garbage.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Make `FormatValidator.ValidateContent` to verify payload of `LOCK`
objects. Pass locked objects to `Locker` interface. Require from
`Locker.Lock` to return `apistatus.IrregularObjectLock` error on a
corresponding condition.
Also add error return to `DeleteHandler.DeleteObjects` method. Require
from method to return `apistatus.ObjectLocked` error on a corresponding
condition. Adopt implementations.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
`Inhume` operation can potentially mark lockers as garbage. There is a
need to update locker list in locked bucket.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Make `StorageEngine.Delete` to forward first encountered
`apistatus.ObjectLocked` error during shard processing.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Make `StorageEngine.Inhume` to forward first encountered
`apistatus.ObjectLocked` error during shard processing.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Implement `StorageEngine.Lock` method which works similar to `Inhume`
but calls `Lock` on the processing shards.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Make `DB.Lock` to return `apistatus.IrregularObjectLock` if at least one
of the locked objects is irregular (not of type REGULAR).
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Make `DB.Inhume` to return `apistatus.ObjectLocked` if at least one of
the inhumed objects is locked.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Make `DB.IterateCoveredByTombstones` to not pass locked objects to the
handler. The method is used by GC, therefore it will not consider locked
objects as candidates for deletion even if their tombstone is expired.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Make `DB.IterateExpired` to not pass locked objects to the handler. The
method is used by GC, therefore it will not consider them as candidates
for deletion.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
After introduction of LOCK objects (of type `TypeLock`) complicated
extended its behavior:
* create `lockers` container bucket (LCB) during PUT;
* remove object from LCB during DELETE;
* look up object in LCB during EXISTS;
* get object from LCB during GET;
* list objects from LCB during LIST with cursor;
* select objects from LCB during SELECT with '*'.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Implement `DB.Lock` method which marks list of the objects as locked by
another object. Only regular objects can be locked.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Create class of container buckets with `LOCKED` suffix. Put identifiers
of the objects of type `LOCK` to these buckets during `DB.Put`.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Remove `Object` and `RawObject` types from `pkg/core/object` package.
Use `Object` type from NeoFS SDK Go library everywhere. Avoid using the
deprecated elements.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Similarly to `Get`. Also fix a bug where `ErrNotFound` is returned
instead of `ErrRangeOutOfBounds`.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
Metabase is expected to contain actual information about objects stored
in shard. If the object is present in metabase but is missing from
blobstor, peform an additional attempt to fetch it directly without
consulting metabase. Such a situation is unexpected, so error counter
is increased for the shard which has the object in the metabase. We
don't increase error counter for the shard which has the object in
blobstor, because some garbage can be expected there. In this
implementation there is no overhead for objects which are really
missing, i.e. are not present in any metabase.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
There are certain errors which are not expected during usual node
operation and which tell us that something is wrong with the shard.
To prevent possible data corruption, move shard in read-only mode after
amount of errors exceeded some threshold. By default no actions are performed.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
If pre-existing blobovnicza is initialized, it's size should be updated
even if all buckets are in place.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
It was added back in 2fb379b7 when we had many shard modes. Now we have
only two and comments for constants are rather descriptive.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
We could also ignore errors during evacuate, but this requires
unmarshaling objects first which slowers the process considerably.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
In read-only mode modifying operations are immediately returned with
error and all background operations are suspended.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
Dump contains magic and a list of objects prefixed by object size in bytes.
We can't use proto-marshaled list because this requires having all dump
in memory. Using TAR induces 512 byte overhead for each object which can
be a problem in some cases.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
For some data compression makes little sense, as it is already compressed.
This commit allows to leave such data unchanged based on `Content-Type`
attribute. Currently exact, prefix and suffix matching are supported.
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
Provide shard mode information via `DumpInfo()`. Delete atomic field from
Shard structure since it duplicates new field.
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
Shard's mode was not used in the Node, so added only two modes whose roles
are clear. More modes will be added in the future.
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
Make `flushBigObjects` routine to mark objects which are written to
`BlobStor`. This prevents already flushed objects from being written on
the next iterator tick.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
For fullness estimation of `Blobovnicza` we use number of object stored
in each size bucket. In previous implementation we multiplied the number
by the difference in bucket boundaries. This expression rather
estimated the minimum volume (and for the smallest bucket, the maximum)
of objects in the bucket.
Multiply number of objects by mean bucket size.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
Make `syncFullnessCounter` to accept `bbolt.Tx` argument of Bolt
transaction within which counter should be synchronized. Pass
corresponding transaction during `Init`.
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
According to BoltDB documentation bucket `value is only valid for the
life of the transaction`.
Make `DB.IsSmall` copy value slice in order to prevent potential memory
corruptions (e.g. `runtime.stringtobyteslice` cast).
Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
