lotus/itests/deals_concurrent_test.go
Aarsh Shah d7076778e2
integrate DAG store and CARv2 in deal-making ()
This commit removes badger from the deal-making processes, and
moves to a new architecture with the dagstore as the cental
component on the miner-side, and CARv2s on the client-side.

Every deal that has been handed off to the sealing subsystem becomes
a shard in the dagstore. Shards are mounted via the LotusMount, which
teaches the dagstore how to load the related piece when serving
retrievals.

When the miner starts the Lotus for the first time with this patch,
we will perform a one-time migration of all active deals into the
dagstore. This is a lightweight process, and it consists simply
of registering the shards in the dagstore.

Shards are backed by the unsealed copy of the piece. This is currently
a CARv1. However, the dagstore keeps CARv2 indices for all pieces, so
when it's time to acquire a shard to serve a retrieval, the unsealed
CARv1 is joined with its index (safeguarded by the dagstore), to form
a read-only blockstore, thus taking the place of the monolithic
badger.

Data transfers have been adjusted to interface directly with CARv2 files.
On inbound transfers (client retrievals, miner storage deals), we stream
the received data into a CARv2 ReadWrite blockstore. On outbound transfers
(client storage deals, miner retrievals), we serve the data off a CARv2
ReadOnly blockstore.

Client-side imports are managed by the refactored *imports.Manager
component (when not using IPFS integration). Just like it before, we use
the go-filestore library to avoid duplicating the data from the original
file in the resulting UnixFS DAG (concretely the leaves). However, the
target of those imports are what we call "ref-CARv2s": CARv2 files placed
under the `$LOTUS_PATH/imports` directory, containing the intermediate
nodes in full, and the leaves as positional references to the original file
on disk.

Client-side retrievals are placed into CARv2 files in the location:
`$LOTUS_PATH/retrievals`.

A new set of `Dagstore*` JSON-RPC operations and `lotus-miner dagstore`
subcommands have been introduced on the miner-side to inspect and manage
the dagstore.

Despite moving to a CARv2-backed system, the IPFS integration has been
respected, and it continues to be possible to make storage deals with data
held in an IPFS node, and to perform retrievals directly into an IPFS node.

NOTE: because the "staging" and "client" Badger blockstores are no longer
used, existing imports on the client will be rendered useless. On startup,
Lotus will enumerate all imports and print WARN statements on the log for
each import that needs to be reimported. These log lines contain these
messages:

- import lacks carv2 path; import will not work; please reimport
- import has missing/broken carv2; please reimport

At the end, we will print a "sanity check completed" message indicating
the count of imports found, and how many were deemed broken.

Co-authored-by: Aarsh Shah <aarshkshah1992@gmail.com>
Co-authored-by: Dirk McCormick <dirkmdev@gmail.com>

Co-authored-by: Raúl Kripalani <raul@protocol.ai>
Co-authored-by: Dirk McCormick <dirkmdev@gmail.com>
2021-08-16 23:34:32 +01:00

210 lines
6.5 KiB
Go

package itests
import (
"context"
"fmt"
"sync"
"testing"
"time"
"github.com/filecoin-project/lotus/chain/actors/policy"
"github.com/stretchr/testify/require"
datatransfer "github.com/filecoin-project/go-data-transfer"
"github.com/filecoin-project/go-state-types/abi"
"github.com/filecoin-project/lotus/api"
"github.com/filecoin-project/lotus/itests/kit"
"github.com/filecoin-project/lotus/node"
"github.com/filecoin-project/lotus/node/modules"
"github.com/filecoin-project/lotus/node/modules/dtypes"
"github.com/filecoin-project/lotus/node/repo"
)
// TestDealWithMarketAndMinerNode is running concurrently a number of storage and retrieval deals towards a miner
// architecture where the `mining/sealing/proving` node is a separate process from the `markets` node
func TestDealWithMarketAndMinerNode(t *testing.T) {
if testing.Short() {
t.Skip("skipping test in short mode")
}
t.Skip("skipping due to flakiness: see #6956")
kit.QuietMiningLogs()
oldDelay := policy.GetPreCommitChallengeDelay()
policy.SetPreCommitChallengeDelay(5)
t.Cleanup(func() {
policy.SetPreCommitChallengeDelay(oldDelay)
})
// For these tests where the block time is artificially short, just use
// a deal start epoch that is guaranteed to be far enough in the future
// so that the deal starts sealing in time
startEpoch := abi.ChainEpoch(8 << 10)
runTest := func(t *testing.T, n int, fastRetrieval bool, carExport bool) {
api.RunningNodeType = api.NodeMiner // TODO(anteva): fix me
client, main, market, _ := kit.EnsembleWithMinerAndMarketNodes(t, kit.ThroughRPC())
dh := kit.NewDealHarness(t, client, main, market)
dh.RunConcurrentDeals(kit.RunConcurrentDealsOpts{
N: n,
FastRetrieval: fastRetrieval,
CarExport: carExport,
StartEpoch: startEpoch,
})
}
// this test is expensive because we don't use mock proofs; do a single cycle.
cycles := []int{4}
for _, n := range cycles {
n := n
ns := fmt.Sprintf("%d", n)
t.Run(ns+"-fastretrieval-CAR", func(t *testing.T) { runTest(t, n, true, true) })
t.Run(ns+"-fastretrieval-NoCAR", func(t *testing.T) { runTest(t, n, true, false) })
t.Run(ns+"-stdretrieval-CAR", func(t *testing.T) { runTest(t, n, false, true) })
t.Run(ns+"-stdretrieval-NoCAR", func(t *testing.T) { runTest(t, n, false, false) })
}
}
func TestDealCyclesConcurrent(t *testing.T) {
if testing.Short() {
t.Skip("skipping test in short mode")
}
oldDelay := policy.GetPreCommitChallengeDelay()
policy.SetPreCommitChallengeDelay(5)
t.Cleanup(func() {
policy.SetPreCommitChallengeDelay(oldDelay)
})
kit.QuietMiningLogs()
// For these tests where the block time is artificially short, just use
// a deal start epoch that is guaranteed to be far enough in the future
// so that the deal starts sealing in time
startEpoch := abi.ChainEpoch(2 << 12)
runTest := func(t *testing.T, n int, fastRetrieval bool, carExport bool) {
client, miner, ens := kit.EnsembleMinimal(t, kit.MockProofs())
ens.InterconnectAll().BeginMining(250 * time.Millisecond)
dh := kit.NewDealHarness(t, client, miner, miner)
dh.RunConcurrentDeals(kit.RunConcurrentDealsOpts{
N: n,
FastRetrieval: fastRetrieval,
CarExport: carExport,
StartEpoch: startEpoch,
})
}
// this test is cheap because we use mock proofs, do various cycles
cycles := []int{2, 4, 8, 16}
for _, n := range cycles {
n := n
ns := fmt.Sprintf("%d", n)
t.Run(ns+"-fastretrieval-CAR", func(t *testing.T) { runTest(t, n, true, true) })
t.Run(ns+"-fastretrieval-NoCAR", func(t *testing.T) { runTest(t, n, true, false) })
t.Run(ns+"-stdretrieval-CAR", func(t *testing.T) { runTest(t, n, false, true) })
t.Run(ns+"-stdretrieval-NoCAR", func(t *testing.T) { runTest(t, n, false, false) })
}
}
func TestSimultanenousTransferLimit(t *testing.T) {
if testing.Short() {
t.Skip("skipping test in short mode")
}
kit.QuietMiningLogs()
oldDelay := policy.GetPreCommitChallengeDelay()
policy.SetPreCommitChallengeDelay(5)
t.Cleanup(func() {
policy.SetPreCommitChallengeDelay(oldDelay)
})
// For these tests where the block time is artificially short, just use
// a deal start epoch that is guaranteed to be far enough in the future
// so that the deal starts sealing in time
startEpoch := abi.ChainEpoch(2 << 12)
const (
graphsyncThrottle = 2
concurrency = 20
)
runTest := func(t *testing.T) {
client, miner, ens := kit.EnsembleMinimal(t, kit.MockProofs(), kit.ConstructorOpts(
node.ApplyIf(node.IsType(repo.StorageMiner), node.Override(new(dtypes.StagingGraphsync), modules.StagingGraphsync(graphsyncThrottle))),
node.Override(new(dtypes.Graphsync), modules.Graphsync(graphsyncThrottle)),
))
ens.InterconnectAll().BeginMining(250 * time.Millisecond)
dh := kit.NewDealHarness(t, client, miner, miner)
ctx, cancel := context.WithCancel(context.Background())
du, err := miner.MarketDataTransferUpdates(ctx)
require.NoError(t, err)
var maxOngoing int
var wg sync.WaitGroup
wg.Add(1)
go func() {
defer wg.Done()
ongoing := map[datatransfer.TransferID]struct{}{}
for {
select {
case u := <-du:
t.Logf("%d - %s", u.TransferID, datatransfer.Statuses[u.Status])
if u.Status == datatransfer.Ongoing && u.Transferred > 0 {
ongoing[u.TransferID] = struct{}{}
} else {
delete(ongoing, u.TransferID)
}
if len(ongoing) > maxOngoing {
maxOngoing = len(ongoing)
}
case <-ctx.Done():
return
}
}
}()
t.Logf("running concurrent deals: %d", concurrency)
dh.RunConcurrentDeals(kit.RunConcurrentDealsOpts{
N: concurrency,
FastRetrieval: true,
StartEpoch: startEpoch,
})
t.Logf("all deals finished")
cancel()
wg.Wait()
// The eventing systems across go-data-transfer and go-graphsync
// are racy, and that's why we can't enforce graphsyncThrottle exactly,
// without making this test racy.
//
// Essentially what could happen is that the graphsync layer starts the
// next transfer before the go-data-transfer FSM has the opportunity to
// move the previously completed transfer to the next stage, thus giving
// the appearance that more than graphsyncThrottle transfers are
// in progress.
//
// Concurrency (20) is x10 higher than graphsyncThrottle (2), so if all
// 20 transfers are not happening at once, we know the throttle is
// in effect. Thus we are a little bit lenient here to account for the
// above races and allow up to graphsyncThrottle*2.
require.LessOrEqual(t, maxOngoing, graphsyncThrottle*2)
}
runTest(t)
}