go: Bump to the new version of the tendermint fork

.changelog/2569.bugfix.1.md

@@ -0,0 +1,10 @@
+Use a newer version of the oasis-core tendermint fork
+
+The updated fork has additional changes to tendermint to hopefully
+prevent the node from crashing if the file descriptors available to the
+process get exhausted due to hitting the rlimit.
+
+While no forward progress can be made while the node is re-opening the
+WAL, the node will now flush incoming connections that are in the process
+of handshaking, and retry re-opening the WAL instead of crashing with
+a panic.

go/common/crash/crash.go

@@ -11,7 +11,6 @@ import (
 	flag "github.com/spf13/pflag"
 	"github.com/spf13/viper"
 
-	"github.com/oasislabs/oasis-core/go/common/debug"
 	"github.com/oasislabs/oasis-core/go/common/logging"
 	"github.com/oasislabs/oasis-core/go/common/random"
 	cmdFlags "github.com/oasislabs/oasis-core/go/oasis-node/cmd/common/flags"
@@ -55,7 +54,7 @@ func newDefaultRandomProvider() RandomProvider {
 }
 
 func defaultCrashMethod() {
-	debug.Trap()
+	runtime.Breakpoint()
 }
 
 var crashGlobal *Crasher

go/consensus/api/submission.go

@@ -4,7 +4,7 @@ import (
 	"context"
 	"time"
 
-	"github.com/cenkalti/backoff"
+	"github.com/cenkalti/backoff/v4"
 
 	"github.com/oasislabs/oasis-core/go/common/crypto/signature"
 	"github.com/oasislabs/oasis-core/go/common/errors"

go/go.mod

@@ -2,7 +2,7 @@ module github.com/oasislabs/oasis-core/go
 
 replace (
 	github.com/tendermint/iavl => github.com/oasislabs/iavl v0.12.0-ekiden3
-	github.com/tendermint/tendermint => github.com/oasislabs/tendermint v0.32.8-oasis1
+	github.com/tendermint/tendermint => github.com/oasislabs/tendermint v0.32.8-oasis2
 	golang.org/x/crypto/curve25519 => github.com/oasislabs/ed25519/extra/x25519 v0.0.0-20191022155220-a426dcc8ad5f
 	golang.org/x/crypto/ed25519 => github.com/oasislabs/ed25519 v0.0.0-20191109133925-b197a691e30d
 )
@@ -14,7 +14,7 @@ require (
 	github.com/blevesearch/go-porterstemmer v1.0.2 // indirect
 	github.com/blevesearch/segment v0.0.0-20160915185041-762005e7a34f // indirect
 	github.com/btcsuite/btcd v0.0.0-20190614013741-962a206e94e9 // indirect
-	github.com/cenkalti/backoff v2.1.1+incompatible
+	github.com/cenkalti/backoff/v4 v4.0.0
 	github.com/codahale/hdrhistogram v0.0.0-20161010025455-3a0bb77429bd // indirect
 	github.com/couchbase/vellum v0.0.0-20190610201045-ec7b775d247f // indirect
 	github.com/cznic/b v0.0.0-20181122101859-a26611c4d92d // indirect

go/go.sum

@@ -43,8 +43,8 @@ github.com/btcsuite/goleveldb v0.0.0-20160330041536-7834afc9e8cd/go.mod h1:F+uVa
 github.com/btcsuite/snappy-go v0.0.0-20151229074030-0bdef8d06723/go.mod h1:8woku9dyThutzjeg+3xrA5iCpBRH8XEEg3lh6TiUghc=
 github.com/btcsuite/websocket v0.0.0-20150119174127-31079b680792/go.mod h1:ghJtEyQwv5/p4Mg4C0fgbePVuGr935/5ddU9Z3TmDRY=
 github.com/btcsuite/winsvc v1.0.0/go.mod h1:jsenWakMcC0zFBFurPLEAyrnc/teJEM1O46fmI40EZs=
-github.com/cenkalti/backoff v2.1.1+incompatible h1:tKJnvO2kl0zmb/jA5UKAt4VoEVw1qxKWjE/Bpp46npY=
-github.com/cenkalti/backoff v2.1.1+incompatible/go.mod h1:90ReRw6GdpyfrHakVjL/QHaoyV4aDUVVkXQJJJ3NXXM=
+github.com/cenkalti/backoff/v4 v4.0.0 h1:6VeaLF9aI+MAUQ95106HwWzYZgJJpZ4stumjj6RFYAU=
+github.com/cenkalti/backoff/v4 v4.0.0/go.mod h1:eEew/i+1Q6OrCDZh3WiXYv3+nJwBASZ8Bog/87DQnVg=
 github.com/census-instrumentation/opencensus-proto v0.2.1/go.mod h1:f6KPmirojxKA12rnyqOA5BBL4O983OfeGPqjHWSTneU=
 github.com/cespare/xxhash v1.1.0 h1:a6HrQnmkObjyL+Gs60czilIUGqrzKutQD6XZog3p+ko=
 github.com/cespare/xxhash v1.1.0/go.mod h1:XrSqR1VqqWfGrhpAt58auRo0WTKS1nRRg3ghfAqPWnc=
@@ -352,8 +352,12 @@ github.com/oasislabs/ed25519 v0.0.0-20191122104632-9d9ffc15f526 h1:xKlK+m6tNFucK
 github.com/oasislabs/ed25519 v0.0.0-20191122104632-9d9ffc15f526/go.mod h1:xIpCyrK2ouGA4QBGbiNbkoONrvJ00u9P3QOkXSOAC0c=
 github.com/oasislabs/iavl v0.12.0-ekiden3 h1:8544fXJb57urhAEpTlIwDBdTJukgpPS/FCS/yj14I8E=
 github.com/oasislabs/iavl v0.12.0-ekiden3/go.mod h1:B/tMpl5cg7n42n3xYQTCckJzQezoI75jedkc8FOiOF0=
+github.com/oasislabs/safeopen v0.0.0-20200117113835-6aa648f43ff8 h1:KC7dcrx0WEeyAWGAG+vdJjmIW36PUfw1x/LUnHjLm2E=
+github.com/oasislabs/safeopen v0.0.0-20200117113835-6aa648f43ff8/go.mod h1:ABsG2IHM7bpTRIH3EvQ8CZQEBkzuhLxXFxaYApYMB9Y=
 github.com/oasislabs/tendermint v0.32.8-oasis1 h1:y+RZsI7D6jFqV/OufXA0meDOgn6eJLsXa5a0iyRLIRU=
 github.com/oasislabs/tendermint v0.32.8-oasis1/go.mod h1:SJjyKg9RLf9FUvqo9sJSpcT/COwormN9U05vR70It/A=
+github.com/oasislabs/tendermint v0.32.8-oasis2 h1:PSEUtAp8Rfe/0T7endF6Iqg4p9+pPkSDSx0E83bD2LM=
+github.com/oasislabs/tendermint v0.32.8-oasis2/go.mod h1:uxexUd6P+G+Zg1yACNBycfcaV1dPI985r79I+IXP38w=
 github.com/oklog/ulid v1.3.1/go.mod h1:CirwcVhetQ6Lv90oh/F+FBtV6XMibvdAFo93nm5qn4U=
 github.com/onsi/ginkgo v1.6.0/go.mod h1:lLunBs/Ym6LB5Z9jYTR76FiuTmxDTDusOGeTQH+WWjE=
 github.com/onsi/ginkgo v1.7.0/go.mod h1:lLunBs/Ym6LB5Z9jYTR76FiuTmxDTDusOGeTQH+WWjE=

go/keymanager/client/client.go

@@ -9,7 +9,7 @@ import (
 	"sync"
 	"time"
 
-	"github.com/cenkalti/backoff"
+	"github.com/cenkalti/backoff/v4"
 	"github.com/pkg/errors"
 	"google.golang.org/grpc"
 	"google.golang.org/grpc/balancer/roundrobin"

go/oasis-node/cmd/common/common.go

@@ -6,6 +6,7 @@ import (
 	"io"
 	"os"
 	"path/filepath"
+	"syscall"
 
 	"github.com/spf13/cobra"
 	flag "github.com/spf13/pflag"
@@ -27,8 +28,13 @@ const (
 	// keys.
 	CfgDebugAllowTestKeys = "debug.allow_test_keys"
 
+	// CfgDebugRlimit is the command flag to set RLIMIT_NOFILE on launch.
+	CfgDebugRlimit = "debug.rlimit"
+
 	cfgConfigFile = "config"
 	CfgDataDir    = "datadir"
+
+	badDefaultRlimit = 1024
 )
 
 var (
@@ -37,6 +43,7 @@ var (
 	rootLog = logging.GetLogger("oasis-node")
 
 	debugAllowTestKeysFlag = flag.NewFlagSet("", flag.ContinueOnError)
+	debugRlimitFlag        = flag.NewFlagSet("", flag.ContinueOnError)
 
 	// RootFlags has the flags that are common across all commands.
 	RootFlags = flag.NewFlagSet("", flag.ContinueOnError)
@@ -90,6 +97,7 @@ func Init() error {
 		initDataDir,
 		initLogging,
 		initPublicKeyBlacklist,
+		initRlimit,
 	}
 
 	for _, fn := range initFns {
@@ -111,16 +119,21 @@ func Logger() *logging.Logger {
 func init() {
 	initLoggingFlags()
 
-	debugAllowTestKeysFlag.Bool(CfgDebugAllowTestKeys, false, "Allow test keys (UNSAFE)")
+	debugAllowTestKeysFlag.Bool(CfgDebugAllowTestKeys, false, "allow test keys (UNSAFE)")
 	_ = debugAllowTestKeysFlag.MarkHidden(CfgDebugAllowTestKeys)
 	_ = viper.BindPFlags(debugAllowTestKeysFlag)
 
+	debugRlimitFlag.Uint64(CfgDebugRlimit, 0, "set RLIMIT_NOFILE")
+	_ = debugRlimitFlag.MarkHidden(CfgDebugRlimit)
+	_ = viper.BindPFlags(debugRlimitFlag)
+
 	RootFlags.StringVar(&cfgFile, cfgConfigFile, "", "config file")
 	RootFlags.String(CfgDataDir, "", "data directory")
 	_ = viper.BindPFlags(RootFlags)
 
 	RootFlags.AddFlagSet(loggingFlags)
 	RootFlags.AddFlagSet(debugAllowTestKeysFlag)
+	RootFlags.AddFlagSet(debugRlimitFlag)
 	RootFlags.AddFlagSet(flags.DebugDontBlameOasisFlag)
 }
 
@@ -184,6 +197,33 @@ func initPublicKeyBlacklist() error {
 	return nil
 }
 
+func initRlimit() error {
+	var rlim syscall.Rlimit
+	if err := syscall.Getrlimit(syscall.RLIMIT_NOFILE, &rlim); err != nil {
+		// Log, but don't return the error, this is only used for testing
+		// and to warn the user if it's set too low.
+		rootLog.Warn("failed to query RLIMIT_NOFILE",
+			"err", err,
+		)
+		return nil
+	}
+
+	if rlim.Cur <= badDefaultRlimit {
+		rootLog.Warn("the node user has a very low RLIMIT_NOFILE, consider increasing",
+			"cur", rlim.Cur,
+			"max", rlim.Max,
+		)
+	}
+
+	desiredLimit := viper.GetUint64(CfgDebugRlimit)
+	if !flags.DebugDontBlameOasis() || desiredLimit == 0 {
+		return nil
+	}
+
+	rlim.Cur = desiredLimit
+	return syscall.Setrlimit(syscall.RLIMIT_NOFILE, &rlim)
+}
+
 // GetOutputWriter will create a file if the config string is set,
 // and otherwise return os.Stdout.
 func GetOutputWriter(cmd *cobra.Command, cfg string) (io.WriteCloser, bool, error) {

go/runtime/client/client.go

@@ -7,7 +7,7 @@ import (
 	"sync"
 	"time"
 
-	"github.com/cenkalti/backoff"
+	"github.com/cenkalti/backoff/v4"
 	"github.com/pkg/errors"
 	"google.golang.org/grpc/codes"
 	"google.golang.org/grpc/status"

go/runtime/tagindexer/tagindexer.go

@@ -5,7 +5,7 @@ import (
 	"context"
 	"time"
 
-	"github.com/cenkalti/backoff"
+	"github.com/cenkalti/backoff/v4"
 
 	"github.com/oasislabs/oasis-core/go/common"
 	"github.com/oasislabs/oasis-core/go/common/logging"

go/storage/client/client.go

@@ -8,7 +8,7 @@ import (
 	"math/rand"
 	"time"
 
-	"github.com/cenkalti/backoff"
+	"github.com/cenkalti/backoff/v4"
 	"github.com/pkg/errors"
 	"google.golang.org/grpc/codes"
 	"google.golang.org/grpc/status"

go/worker/common/p2p/p2p.go

@@ -8,7 +8,7 @@ import (
 	"sync"
 	"time"
 
-	"github.com/cenkalti/backoff"
+	"github.com/cenkalti/backoff/v4"
 	"github.com/libp2p/go-libp2p"
 	"github.com/libp2p/go-libp2p-core"
 	"github.com/libp2p/go-libp2p-core/helpers"

go/worker/registration/worker.go

@@ -7,7 +7,7 @@ import (
 	"sync"
 	"sync/atomic"
 
-	"github.com/cenkalti/backoff"
+	"github.com/cenkalti/backoff/v4"
 	"github.com/pkg/errors"
 	flag "github.com/spf13/pflag"
 	"github.com/spf13/viper"