da9ca0c650
Satellites set their configuration values to default values using cfgstruct, however, it turns out our tests don't test these values at all! Instead, they have a completely separate definition system that is easy to forget about. As is to be expected, these values have drifted, and it appears in a few cases test planet is testing unreasonable values that we won't see in production, or perhaps worse, features enabled in production were missed and weren't enabled in testplanet. This change makes it so all values are configured the same, systematic way, so it's easy to see when test values are different than dev values or release values, and it's less hard to forget to enable features in testplanet. In terms of reviewing, this change should be actually fairly easy to review, considering private/testplanet/satellite.go keeps the current config system and the new one and confirms that they result in identical configurations, so you can be certain that nothing was missed and the config is all correct. You can also check the config lock to see what actual config values changed. Change-Id: I6715d0794887f577e21742afcf56fd2b9d12170e
186 lines
5.7 KiB
Go
186 lines
5.7 KiB
Go
// Copyright (C) 2019 Storj Labs, Inc.
|
|
// See LICENSE for copying information.
|
|
|
|
package rollup
|
|
|
|
import (
|
|
"context"
|
|
"time"
|
|
|
|
"go.uber.org/zap"
|
|
|
|
"storj.io/common/pb"
|
|
"storj.io/common/storj"
|
|
"storj.io/common/sync2"
|
|
"storj.io/storj/satellite/accounting"
|
|
)
|
|
|
|
// Config contains configurable values for rollup.
|
|
type Config struct {
|
|
Interval time.Duration `help:"how frequently rollup should run" releaseDefault:"24h" devDefault:"120s" testDefault:"$TESTINTERVAL"`
|
|
DeleteTallies bool `help:"option for deleting tallies after they are rolled up" default:"true"`
|
|
}
|
|
|
|
// Service is the rollup service for totalling data on storage nodes on daily intervals.
|
|
//
|
|
// architecture: Chore
|
|
type Service struct {
|
|
logger *zap.Logger
|
|
Loop *sync2.Cycle
|
|
sdb accounting.StoragenodeAccounting
|
|
deleteTallies bool
|
|
OrderExpiration time.Duration
|
|
}
|
|
|
|
// New creates a new rollup service.
|
|
func New(logger *zap.Logger, sdb accounting.StoragenodeAccounting, interval time.Duration, deleteTallies bool, orderExpiration time.Duration) *Service {
|
|
return &Service{
|
|
logger: logger,
|
|
Loop: sync2.NewCycle(interval),
|
|
sdb: sdb,
|
|
deleteTallies: deleteTallies,
|
|
OrderExpiration: orderExpiration,
|
|
}
|
|
}
|
|
|
|
// Run the Rollup loop.
|
|
func (r *Service) Run(ctx context.Context) (err error) {
|
|
defer mon.Task()(&ctx)(&err)
|
|
return r.Loop.Run(ctx, func(ctx context.Context) error {
|
|
err := r.Rollup(ctx)
|
|
if err != nil {
|
|
r.logger.Error("rollup failed", zap.Error(err))
|
|
}
|
|
return nil
|
|
})
|
|
}
|
|
|
|
// Close stops the service and releases any resources.
|
|
func (r *Service) Close() error {
|
|
r.Loop.Close()
|
|
return nil
|
|
}
|
|
|
|
// Rollup aggregates storage and bandwidth amounts for the time interval.
|
|
func (r *Service) Rollup(ctx context.Context) (err error) {
|
|
defer mon.Task()(&ctx)(&err)
|
|
// only Rollup new things - get LastRollup
|
|
lastRollup, err := r.sdb.LastTimestamp(ctx, accounting.LastRollup)
|
|
if err != nil {
|
|
return Error.Wrap(err)
|
|
}
|
|
// unexpired orders with created at times before the last rollup timestamp could still have been added later
|
|
if !lastRollup.IsZero() {
|
|
lastRollup = lastRollup.Add(-r.OrderExpiration)
|
|
}
|
|
|
|
rollupStats := make(accounting.RollupStats)
|
|
latestTally, err := r.RollupStorage(ctx, lastRollup, rollupStats)
|
|
if err != nil {
|
|
return Error.Wrap(err)
|
|
}
|
|
|
|
err = r.RollupBW(ctx, lastRollup, rollupStats)
|
|
if err != nil {
|
|
return Error.Wrap(err)
|
|
}
|
|
|
|
// remove the latest day (which we cannot know is complete), then push to DB
|
|
latestTally = time.Date(latestTally.Year(), latestTally.Month(), latestTally.Day(), 0, 0, 0, 0, latestTally.Location())
|
|
delete(rollupStats, latestTally)
|
|
if len(rollupStats) == 0 {
|
|
r.logger.Info("RollupStats is empty")
|
|
return nil
|
|
}
|
|
|
|
err = r.sdb.SaveRollup(ctx, latestTally, rollupStats)
|
|
if err != nil {
|
|
return Error.Wrap(err)
|
|
}
|
|
|
|
if r.deleteTallies {
|
|
// Delete already rolled up tallies
|
|
latestTally = latestTally.Add(-r.OrderExpiration)
|
|
err = r.sdb.DeleteTalliesBefore(ctx, latestTally)
|
|
if err != nil {
|
|
return Error.Wrap(err)
|
|
}
|
|
}
|
|
|
|
return nil
|
|
}
|
|
|
|
// RollupStorage rolls up storage tally, modifies rollupStats map.
|
|
func (r *Service) RollupStorage(ctx context.Context, lastRollup time.Time, rollupStats accounting.RollupStats) (latestTally time.Time, err error) {
|
|
defer mon.Task()(&ctx)(&err)
|
|
tallies, err := r.sdb.GetTalliesSince(ctx, lastRollup)
|
|
if err != nil {
|
|
return lastRollup, Error.Wrap(err)
|
|
}
|
|
if len(tallies) == 0 {
|
|
r.logger.Info("Rollup found no new tallies")
|
|
return lastRollup, nil
|
|
}
|
|
// loop through tallies and build Rollup
|
|
for _, tallyRow := range tallies {
|
|
node := tallyRow.NodeID
|
|
// tallyEndTime is the time the at rest tally was saved
|
|
tallyEndTime := tallyRow.IntervalEndTime.UTC()
|
|
if tallyEndTime.After(latestTally) {
|
|
latestTally = tallyEndTime
|
|
}
|
|
// create or get AccoutingRollup day entry
|
|
iDay := time.Date(tallyEndTime.Year(), tallyEndTime.Month(), tallyEndTime.Day(), 0, 0, 0, 0, tallyEndTime.Location())
|
|
if rollupStats[iDay] == nil {
|
|
rollupStats[iDay] = make(map[storj.NodeID]*accounting.Rollup)
|
|
}
|
|
if rollupStats[iDay][node] == nil {
|
|
rollupStats[iDay][node] = &accounting.Rollup{NodeID: node, StartTime: iDay}
|
|
}
|
|
// increment data at rest sum
|
|
rollupStats[iDay][node].AtRestTotal += tallyRow.DataTotal
|
|
}
|
|
|
|
return latestTally, nil
|
|
}
|
|
|
|
// RollupBW aggregates the bandwidth rollups, modifies rollupStats map.
|
|
func (r *Service) RollupBW(ctx context.Context, lastRollup time.Time, rollupStats accounting.RollupStats) (err error) {
|
|
defer mon.Task()(&ctx)(&err)
|
|
err = r.sdb.GetBandwidthSince(ctx, lastRollup.UTC(), func(ctx context.Context, row *accounting.StoragenodeBandwidthRollup) error {
|
|
nodeID := row.NodeID
|
|
// interval is the time the bw order was saved
|
|
interval := row.IntervalStart.UTC()
|
|
day := time.Date(interval.Year(), interval.Month(), interval.Day(), 0, 0, 0, 0, interval.Location())
|
|
if rollupStats[day] == nil {
|
|
rollupStats[day] = make(map[storj.NodeID]*accounting.Rollup)
|
|
}
|
|
if rollupStats[day][nodeID] == nil {
|
|
rollupStats[day][nodeID] = &accounting.Rollup{NodeID: nodeID, StartTime: day}
|
|
}
|
|
switch row.Action {
|
|
case uint(pb.PieceAction_INVALID):
|
|
r.logger.Info("invalid order action type")
|
|
case uint(pb.PieceAction_PUT):
|
|
rollupStats[day][nodeID].PutTotal += int64(row.Settled)
|
|
case uint(pb.PieceAction_GET):
|
|
rollupStats[day][nodeID].GetTotal += int64(row.Settled)
|
|
case uint(pb.PieceAction_GET_AUDIT):
|
|
rollupStats[day][nodeID].GetAuditTotal += int64(row.Settled)
|
|
case uint(pb.PieceAction_GET_REPAIR):
|
|
rollupStats[day][nodeID].GetRepairTotal += int64(row.Settled)
|
|
case uint(pb.PieceAction_PUT_REPAIR):
|
|
rollupStats[day][nodeID].PutRepairTotal += int64(row.Settled)
|
|
default:
|
|
r.logger.Info("delete order type")
|
|
}
|
|
|
|
return nil
|
|
})
|
|
if err != nil {
|
|
return Error.Wrap(err)
|
|
}
|
|
|
|
return nil
|
|
}
|