storj/satellite/accounting/db.go

256 lines
11 KiB
Go
Raw Normal View History

2019-01-24 20:15:10 +00:00
// Copyright (C) 2019 Storj Labs, Inc.
// See LICENSE for copying information.
package accounting
import (
"context"
"time"
"storj.io/common/memory"
"storj.io/common/storj"
"storj.io/common/uuid"
"storj.io/storj/satellite/compensation"
"storj.io/storj/satellite/metabase"
"storj.io/storj/satellite/orders"
)
// RollupStats is a convenience alias.
type RollupStats map[time.Time]map[storj.NodeID]*Rollup
// StoragenodeStorageTally mirrors dbx.StoragenodeStorageTally, allowing us to use that struct without leaking dbx.
type StoragenodeStorageTally struct {
ID int64
NodeID storj.NodeID
IntervalEndTime time.Time
DataTotal float64
}
// StoragenodeBandwidthRollup mirrors dbx.StoragenodeBandwidthRollup, allowing us to use the struct without leaking dbx.
type StoragenodeBandwidthRollup struct {
NodeID storj.NodeID
IntervalStart time.Time
Action uint
Settled uint64
}
// Rollup mirrors dbx.AccountingRollup, allowing us to use that struct without leaking dbx.
type Rollup struct {
ID int64
NodeID storj.NodeID
StartTime time.Time
PutTotal int64
GetTotal int64
GetAuditTotal int64
GetRepairTotal int64
PutRepairTotal int64
AtRestTotal float64
}
// StorageNodePeriodUsage represents a statement for a node for a compensation period.
type StorageNodePeriodUsage struct {
NodeID storj.NodeID
AtRestTotal float64
GetTotal int64
PutTotal int64
GetRepairTotal int64
PutRepairTotal int64
GetAuditTotal int64
}
// StorageNodeUsage is node at rest space usage over a period of time.
type StorageNodeUsage struct {
NodeID storj.NodeID
StorageUsed float64
Timestamp time.Time
}
// ProjectUsage consist of period total storage, egress
// and objects count per hour for certain Project in bytes.
type ProjectUsage struct {
Storage float64 `json:"storage"`
Egress int64 `json:"egress"`
ObjectCount float64 `json:"objectCount"`
Since time.Time `json:"since"`
Before time.Time `json:"before"`
}
satellite/accounting: add cache for getting project storage and bw limits This PR adds the following items: 1) an in-memory read-only cache thats stores project limit info for projectIDs This cache is stored in-memory since this is expected to be a small amount of data. In this implementation we are only storing in the cache projects that have been accessed. Currently for the largest Satellite (eu-west) there is about 4500 total projects. So storing the storage limit (int64) and the bandwidth limit (int64), this would end up being about 200kb (including the 32 byte project ID) if all 4500 projectIDs were in the cache. So this all fits in memory for the time being. At some point it may not as usage grows, but that seems years out. The cache is a read only cache. When requests come in to upload/download a file, we will read from the cache what the current limits are for that project. If the cache does not contain the projectID, it will get the info from the database (satellitedb project table), then add it to the cache. The only time the values in the cache are modified is when either a) the project ID is not in the cache, or b) the item in the cache has expired (default 10mins), then the data gets refreshed out of the database. This occurs by default every 10 mins. This means that if we update the usage limits in the database, that change might not show up in the cache for 10 mins which mean it will not be reflected to limit end users uploading/downloading files for that time period.. Change-Id: I3fd7056cf963676009834fcbcf9c4a0922ca4a8f
2020-09-09 20:20:44 +01:00
// ProjectLimits contains the storage and bandwidth limits.
type ProjectLimits struct {
Usage *int64
Bandwidth *int64
}
// BucketUsage consist of total bucket usage for period.
type BucketUsage struct {
ProjectID uuid.UUID
BucketName string
Storage float64
Egress float64
ObjectCount int64
Since time.Time
Before time.Time
}
// BucketUsageCursor holds info for bucket usage
// cursor pagination.
type BucketUsageCursor struct {
Search string
Limit uint
Page uint
}
// BucketUsagePage represents bucket usage page result.
type BucketUsagePage struct {
BucketUsages []BucketUsage
Search string
Limit uint
Offset uint64
PageCount uint
CurrentPage uint
TotalCount uint64
}
// BucketUsageRollup is total bucket usage info
// for certain period.
type BucketUsageRollup struct {
ProjectID uuid.UUID
BucketName []byte
RemoteStoredData float64
InlineStoredData float64
RemoteSegments float64
InlineSegments float64
ObjectCount float64
MetadataSize float64
RepairEgress float64
GetEgress float64
AuditEgress float64
Since time.Time
Before time.Time
}
// StoragenodeAccounting stores information about bandwidth and storage usage for storage nodes.
2019-09-10 14:24:16 +01:00
//
// architecture: Database
type StoragenodeAccounting interface {
// SaveTallies records tallies of data at rest
SaveTallies(ctx context.Context, latestTally time.Time, nodeData map[storj.NodeID]float64) error
// GetTallies retrieves all tallies
GetTallies(ctx context.Context) ([]*StoragenodeStorageTally, error)
// GetTalliesSince retrieves all tallies since latestRollup
GetTalliesSince(ctx context.Context, latestRollup time.Time) ([]*StoragenodeStorageTally, error)
// GetBandwidthSince retrieves all bandwidth rollup entires since latestRollup
GetBandwidthSince(ctx context.Context, latestRollup time.Time, cb func(context.Context, *StoragenodeBandwidthRollup) error) error
// SaveRollup records tally and bandwidth rollup aggregations to the database
SaveRollup(ctx context.Context, latestTally time.Time, stats RollupStats) error
// LastTimestamp records and returns the latest last tallied time.
LastTimestamp(ctx context.Context, timestampType string) (time.Time, error)
// QueryPaymentInfo queries Nodes and Accounting_Rollup on nodeID
2019-01-17 18:34:13 +00:00
QueryPaymentInfo(ctx context.Context, start time.Time, end time.Time) ([]*CSVRow, error)
// QueryStorageNodePeriodUsage returns accounting statements for nodes for a given compensation period
QueryStorageNodePeriodUsage(ctx context.Context, period compensation.Period) ([]StorageNodePeriodUsage, error)
// QueryStorageNodeUsage returns slice of StorageNodeUsage for given period
QueryStorageNodeUsage(ctx context.Context, nodeID storj.NodeID, start time.Time, end time.Time) ([]StorageNodeUsage, error)
// DeleteTalliesBefore deletes all tallies prior to some time
DeleteTalliesBefore(ctx context.Context, latestRollup time.Time) error
// ArchiveRollupsBefore archives rollups older than a given time and returns num storagenode and bucket bandwidth rollups archived.
ArchiveRollupsBefore(ctx context.Context, before time.Time, batchSize int) (numArchivedNodeBW int, err error)
// GetRollupsSince retrieves all archived bandwidth rollup records since a given time. A hard limit batch size is used for results.
GetRollupsSince(ctx context.Context, since time.Time) ([]StoragenodeBandwidthRollup, error)
// GetArchivedRollupsSince retrieves all archived bandwidth rollup records since a given time. A hard limit batch size is used for results.
GetArchivedRollupsSince(ctx context.Context, since time.Time) ([]StoragenodeBandwidthRollup, error)
}
// ProjectAccounting stores information about bandwidth and storage usage for projects.
2019-09-10 14:24:16 +01:00
//
// architecture: Database
type ProjectAccounting interface {
// SaveTallies saves the latest project info
SaveTallies(ctx context.Context, intervalStart time.Time, bucketTallies map[metabase.BucketLocation]*BucketTally) error
// GetTallies retrieves all tallies
GetTallies(ctx context.Context) ([]BucketTally, error)
// CreateStorageTally creates a record for BucketStorageTally in the accounting DB table
CreateStorageTally(ctx context.Context, tally BucketStorageTally) error
// GetAllocatedBandwidthTotal returns the sum of GET bandwidth usage allocated for a projectID in the past time frame
GetAllocatedBandwidthTotal(ctx context.Context, projectID uuid.UUID, from time.Time) (int64, error)
// GetProjectAllocatedBandwidth returns project allocated bandwidth for the specified year and month.
GetProjectAllocatedBandwidth(ctx context.Context, projectID uuid.UUID, year int, month time.Month) (int64, error)
// DeleteProjectAllocatedBandwidthBefore deletes project bandwidth rollups before the given time
DeleteProjectAllocatedBandwidthBefore(ctx context.Context, before time.Time) error
// GetStorageTotals returns the current inline and remote storage usage for a projectID
GetStorageTotals(ctx context.Context, projectID uuid.UUID) (int64, int64, error)
// UpdateProjectUsageLimit updates project usage limit.
UpdateProjectUsageLimit(ctx context.Context, projectID uuid.UUID, limit memory.Size) error
// UpdateProjectBandwidthLimit updates project bandwidth limit.
UpdateProjectBandwidthLimit(ctx context.Context, projectID uuid.UUID, limit memory.Size) error
// GetProjectStorageLimit returns project storage usage limit.
GetProjectStorageLimit(ctx context.Context, projectID uuid.UUID) (*int64, error)
// GetProjectBandwidthLimit returns project bandwidth usage limit.
GetProjectBandwidthLimit(ctx context.Context, projectID uuid.UUID) (*int64, error)
satellite/accounting: add cache for getting project storage and bw limits This PR adds the following items: 1) an in-memory read-only cache thats stores project limit info for projectIDs This cache is stored in-memory since this is expected to be a small amount of data. In this implementation we are only storing in the cache projects that have been accessed. Currently for the largest Satellite (eu-west) there is about 4500 total projects. So storing the storage limit (int64) and the bandwidth limit (int64), this would end up being about 200kb (including the 32 byte project ID) if all 4500 projectIDs were in the cache. So this all fits in memory for the time being. At some point it may not as usage grows, but that seems years out. The cache is a read only cache. When requests come in to upload/download a file, we will read from the cache what the current limits are for that project. If the cache does not contain the projectID, it will get the info from the database (satellitedb project table), then add it to the cache. The only time the values in the cache are modified is when either a) the project ID is not in the cache, or b) the item in the cache has expired (default 10mins), then the data gets refreshed out of the database. This occurs by default every 10 mins. This means that if we update the usage limits in the database, that change might not show up in the cache for 10 mins which mean it will not be reflected to limit end users uploading/downloading files for that time period.. Change-Id: I3fd7056cf963676009834fcbcf9c4a0922ca4a8f
2020-09-09 20:20:44 +01:00
// GetProjectLimits returns current project limit for both storage and bandwidth.
GetProjectLimits(ctx context.Context, projectID uuid.UUID) (ProjectLimits, error)
// GetProjectTotal returns project usage summary for specified period of time.
GetProjectTotal(ctx context.Context, projectID uuid.UUID, since, before time.Time) (*ProjectUsage, error)
// GetBucketUsageRollups returns usage rollup per each bucket for specified period of time.
GetBucketUsageRollups(ctx context.Context, projectID uuid.UUID, since, before time.Time) ([]BucketUsageRollup, error)
// GetBucketTotals returns per bucket usage summary for specified period of time.
GetBucketTotals(ctx context.Context, projectID uuid.UUID, cursor BucketUsageCursor, since, before time.Time) (*BucketUsagePage, error)
// ArchiveRollupsBefore archives rollups older than a given time and returns number of bucket bandwidth rollups archived.
ArchiveRollupsBefore(ctx context.Context, before time.Time, batchSize int) (numArchivedBucketBW int, err error)
// GetRollupsSince retrieves all archived bandwidth rollup records since a given time. A hard limit batch size is used for results.
GetRollupsSince(ctx context.Context, since time.Time) ([]orders.BucketBandwidthRollup, error)
// GetArchivedRollupsSince retrieves all archived bandwidth rollup records since a given time. A hard limit batch size is used for results.
GetArchivedRollupsSince(ctx context.Context, since time.Time) ([]orders.BucketBandwidthRollup, error)
}
// Cache stores live information about project storage which has not yet been synced to ProjectAccounting.
//
// All the implementations must follow the convention of returning errors of one
// of the classes defined in this package.
//
// All the methods return:
//
// ErrInvalidArgument: an implementation may return if some parameter contain a
// value which isn't accepted, nonetheless, not all the implementations impose
// the same constraints on them.
//
// ErrSystemOrNetError: any method will return this if there is an error with
// the underlining system or the network.
//
// ErrKeyNotFound: returned when a key is not found.
//
// ErrUnexpectedValue: returned when a key or value stored in the underlying
// system isn't of the expected format or type according the business domain.
//
// architecture: Database
type Cache interface {
// GetProjectStorageUsage returns the project's storage usage.
GetProjectStorageUsage(ctx context.Context, projectID uuid.UUID) (totalUsed int64, err error)
// GetProjectBandwidthUsage returns the project's bandwidth usage.
GetProjectBandwidthUsage(ctx context.Context, projectID uuid.UUID, now time.Time) (currentUsed int64, err error)
// UpdateProjectBandthUsage updates the project's bandwidth usage increasing
// it. The projectID is inserted to the increment when it doesn't exists,
// hence this method will never return ErrKeyNotFound error's class.
UpdateProjectBandwidthUsage(ctx context.Context, projectID uuid.UUID, increment int64, ttl time.Duration, now time.Time) error
// AddProjectStorageUsage adds to the projects storage usage the spacedUsed.
// The projectID is inserted to the spaceUsed when it doesn't exists, hence
// this method will never return ErrKeyNotFound.
satellite/accounting: refactor live accounting to hold current estimated totals live accounting used to be a cache to store writes before they are picked up during the tally iteration, after which the cache is cleared. This created a window in which users could potentially exceed the storage limit. This PR refactors live accounting to hold current estimations of space used per project. This should also reduce DB load since we no longer need to query the satellite DB when checking space used for limiting. The mechanism by which the new live accounting system works is as follows: During the upload of any segment, the size of that segment is added to its respective project total in live accounting. At the beginning of the tally iteration we record the current values in live accounting as `initialLiveTotals`. At the end of the tally iteration we again record the current totals in live accounting as `latestLiveTotals`. The metainfo loop observer in tally allows us to get the project totals from what it observed in metainfo DB which are stored in `tallyProjectTotals`. However, for any particular segment uploaded during the metainfo loop, the observer may or may not have seen it. Thus, we take half of the difference between `latestLiveTotals` and `initialLiveTotals`, and add that to the total that was found during tally and set that as the new live accounting total. Initially, live accounting was storing the total stored amount across all nodes rather than the segment size, which is inconsistent with how we record amounts stored in the project accounting DB, so we have refactored live accounting to record segment size Change-Id: Ie48bfdef453428fcdc180b2d781a69d58fd927fb
2019-10-31 17:27:38 +00:00
AddProjectStorageUsage(ctx context.Context, projectID uuid.UUID, spaceUsed int64) error
// GetAllProjectTotals return the total projects' storage used space.
satellite/accounting: refactor live accounting to hold current estimated totals live accounting used to be a cache to store writes before they are picked up during the tally iteration, after which the cache is cleared. This created a window in which users could potentially exceed the storage limit. This PR refactors live accounting to hold current estimations of space used per project. This should also reduce DB load since we no longer need to query the satellite DB when checking space used for limiting. The mechanism by which the new live accounting system works is as follows: During the upload of any segment, the size of that segment is added to its respective project total in live accounting. At the beginning of the tally iteration we record the current values in live accounting as `initialLiveTotals`. At the end of the tally iteration we again record the current totals in live accounting as `latestLiveTotals`. The metainfo loop observer in tally allows us to get the project totals from what it observed in metainfo DB which are stored in `tallyProjectTotals`. However, for any particular segment uploaded during the metainfo loop, the observer may or may not have seen it. Thus, we take half of the difference between `latestLiveTotals` and `initialLiveTotals`, and add that to the total that was found during tally and set that as the new live accounting total. Initially, live accounting was storing the total stored amount across all nodes rather than the segment size, which is inconsistent with how we record amounts stored in the project accounting DB, so we have refactored live accounting to record segment size Change-Id: Ie48bfdef453428fcdc180b2d781a69d58fd927fb
2019-10-31 17:27:38 +00:00
GetAllProjectTotals(ctx context.Context) (map[uuid.UUID]int64, error)
// Close the client, releasing any open resources. Once it's called any other
// method must be called.
Close() error
}