feat(executor): added MsgUpsertSequencer consensus message #1120

keruch · 2024-10-04T17:43:00Z

PR Standards

ADR link https://www.notion.so/dymension/ADR-x-Sequencer-Reward-Address-On-Rollapp-da84af6888c141d0a4c1a8df256a5025
Closes dymensionxyz/dymension#1248

For Author:

Targeted PR against correct branch
included the correct type prefix in the PR title
Linked to Github issue with discussion and accepted design
Targets only one github issue
Wrote unit and integration tests
All CI checks have passed
Added relevant godoc comments

For Reviewer:

confirmed the correct type prefix in the PR title
Reviewers assigned
confirmed all author checklist items have been addressed

After reviewer approval:

In case targets main branch, PR should be squashed and merged.
In case PR targets a release branch, PR should be rebased.

block/produce.go

…ard-addr

faultytolly · 2024-10-09T20:47:42Z

I think there should be a way to test it

zale144 · 2024-10-09T20:36:45Z

block/produce.go

+		return []proto.Message{&sequencertypes.MsgUpsertSequencer{
+			Operator:        nextProposerSettlementAddr,
+			ConsPubKey:      anyPubKey,
+			RewardAddrBytes: addrBytes,


For the initial RewardAddress, the settlementAddr is used. Do we need to include both in the consensus message? Is there a scenario where this message can carry two distinct addresses, one Operator and the other Reward address?
Do I understand this correctly, to have a reward address different from the settlement addr, the custom reward address can be set on the rdk manually, in which case the RewardAddrBytes will not override it?
I realize keeping the reward address in the Hub was decided against, but if we recognize the Hub's x/sequencer record to be the source of truth for the sequencer, then we could maybe consider that besides the settlement address, the (custom) reward address could also originate from the Hub's sequencer record. What do you think?

Regarding your last point. Yes, you are right. That's what we finally decided to do with Omri. We'll enrich the rotation_started event with all the information needed for the sequencer, ie RewardAddr and WhitelistedRelayers so far. And the hub will be the source of truth.

Do we need to include both in the consensus message?

Good point, probably we don't. But as i said above, this will be change anyway soon.

Do I understand this correctly, to have a reward address different from the settlement addr, the custom reward address can be set on the rdk manually, in which case the RewardAddrBytes will not override it?

Currently, RewardAddrBytes overwrites the manually set value. Again, this will change once we accept that the Hub is the source of truce.

keruch · 2024-10-10T07:51:00Z

I think there should be a way to test it

@faultytolly Existing tests already cover the code I added. I even adjusted some existing scenarios to make tests green, you can see the diff.

keruch · 2024-10-10T07:52:11Z

The build fails while the RDK PR dymensionxyz/dymension-rdk#568 is not merged.

omritoptix

looks good. small code cleanliness comments.

omritoptix · 2024-10-12T09:43:06Z

block/produce.go

+	nextProposerSettlementAddr string,
+	lastSeqBlock bool, // Indicates that the block is the last for the current seq. True during the rotation.
+) ([]proto.Message, error) {
+	if m.State.IsGenesis() || lastSeqBlock {


for code cleanliness we can improve it by not having all the code under the if, just return fast if it's not. (i.e if !m.state.isGenesis() && !lastSeqBlock -> return)

omritoptix

I notice now we introduce a circular dependency with the RDK which makes this pr fails to build also.

I think we should avoid the circular depedency, it makes it really difiicult to maintain after.

can't we just copy the protos?

block/produce.go

keruch · 2024-10-14T22:17:25Z

@omritoptix deleted a circular dependency with the RDK

mtsitrin · 2024-10-15T09:03:48Z

p2p/client_test.go

@@ -49,6 +50,7 @@ func TestClientStartup(t *testing.T) {
 	assert.NoError(err)
 }

+// TestBootstrapping TODO: this test is flaky in main. Try running it 100 times to reproduce.


add reference to #869

mtsitrin · 2024-10-15T09:18:19Z

block/produce.go

@@ -96,18 +100,26 @@ func (m *Manager) ProduceBlockLoop(ctx context.Context, bytesProducedC chan int)
 	}
 }

+// nextProposerInfo holds information about the next proposer.
+type nextProposerInfo struct {


this nextProposerInfo struct is uneeded, and changes alot of APIs just to be used in consensusMsgsOnCreateBlock.
nextProposerAddr can be queried using m.State.Sequencers.GetByHash (similar to SetByHash)

mtsitrin · 2024-10-15T09:19:52Z

block/produce.go

+// Currently, we need to create a sequencer in the rollapp if it doesn't exist in the following cases:
+//   - On the very first block after the genesis or
+//   - On the last block of the current sequencer (eg, during the rotation).
+func (m *Manager) consensusMsgsOnCreateBlock(


i suggest to move to separate file, to have all consesnsMsg related logic

mtsitrin · 2024-10-15T09:22:47Z

block/produce.go

 	tmed25519 "github.com/tendermint/tendermint/crypto/ed25519"
 	cmtproto "github.com/tendermint/tendermint/proto/tendermint/types"
 	tmtypes "github.com/tendermint/tendermint/types"
 	tmtime "github.com/tendermint/tendermint/types/time"

+	sequencertypes "github.com/dymensionxyz/dymint/types/pb/rollapp/sequencers/types"


Suggested change

sequencertypes "github.com/dymensionxyz/dymint/types/pb/rollapp/sequencers/types"

rdktypes "github.com/dymensionxyz/dymint/types/pb/rollapp/sequencers/types"

mtsitrin · 2024-10-15T09:24:41Z

block/produce.go

+	}
+
+	return []proto.Message{&sequencertypes.ConsensusMsgUpsertSequencer{
+		Operator:   nextProposerSettlementAddr,


we're passing "dym1...." to the RDK, which needs to be converted to the rollapp's native prefix.
Maybe we should pass the []byte of the address, not the bech32 representation

i had bytes here initially but i thought it wasn't convenient, so i added a method in sdk-utils to convert an arbitrary-prefixed bech32 string to AccAddress and reused it everywhere in rdk

func FromBech32[T sdk.AccAddress | sdk.ValAddress](addr string) (T, error)

dymensionxyz/sdk-utils#11

Faulty Tolly and others added 15 commits September 19, 2024 16:08

temp commit

ca7a5a4

temp commit

e382846

add consensus messages to the block data

c6639e7

Execute block

336e3ca

previous proto gen

9209ab6

add consensus messages stream

d08e81b

add some fixes

45164f1

goimports

adb695c

remove nil check

f24ea2f

goimports

0fa4c8d

remove line from proto

4b061ec

remove proto comment

038a2a3

Merge branch 'main' into feat/consensus-messages

130d86c

merge commit

102fa8e

feat(executor): added MsgUpsertSequencer consensus message

dc2cafe

keruch self-assigned this Oct 4, 2024

keruch mentioned this pull request Oct 4, 2024

Sequencer Reward Address On Rollapp - Impl dymensionxyz/dymension#1248

Closed

github-actions bot added the dym-internal label Oct 4, 2024

keruch added 3 commits October 7, 2024 17:31

feat(executor): MsgUpsertSequecner

9f507a4

tests fix 1

d55070b

tests fix 2

ef2a708

keruch changed the base branch from main to feat/consensus-messages October 7, 2024 17:12

keruch changed the title ~~Kirill/1248 sequencer reward addr~~ feat(executor): added MsgUpsertSequencer consensus message Oct 7, 2024

keruch marked this pull request as ready for review October 7, 2024 17:13

keruch requested a review from a team as a code owner October 7, 2024 17:13

Updated cometbft version.

f357f7d

faultytolly reviewed Oct 9, 2024

View reviewed changes

block/produce.go Show resolved Hide resolved

Merge branch 'feat/consensus-messages' into kirill/1248-sequencer-rew…

08965fc

…ard-addr

Base automatically changed from feat/consensus-messages to main October 9, 2024 13:04

Merge branch 'main' into kirill/1248-sequencer-reward-addr

0f93728

keruch mentioned this pull request Oct 9, 2024

Implement ConsensusMessagesStream #1125

Open

linked tasks

ae84159

faultytolly previously approved these changes Oct 9, 2024

View reviewed changes

zale144 reviewed Oct 9, 2024

View reviewed changes

omritoptix approved these changes Oct 12, 2024

View reviewed changes

omritoptix requested changes Oct 12, 2024

View reviewed changes

keruch added 2 commits October 14, 2024 20:50

Merge branch 'main' into kirill/1248-sequencer-reward-addr

6b340dd

feat(proto): added hub and rdk x/sequencer

ab8d8d2

keruch dismissed faultytolly’s stale review via ab8d8d2 October 14, 2024 22:07

github-advanced-security bot found potential problems Oct 14, 2024

View reviewed changes

block/produce.go Dismissed Show dismissed Hide dismissed

keruch requested a review from omritoptix October 14, 2024 22:23

keruch added 2 commits October 15, 2024 00:35

feat(proto): added hub and rdk x/sequencer

fae24da

Merge branch 'kirill/1083-proto' into kirill/1248-sequencer-reward-addr

cffbeab

keruch changed the base branch from main to kirill/1083-proto October 14, 2024 22:42

fix build

05b4930

mtsitrin requested changes Oct 15, 2024

View reviewed changes

Base automatically changed from kirill/1083-proto to main October 15, 2024 10:22

keruch added 3 commits October 15, 2024 20:33

Merge branch 'main' into kirill/1248-sequencer-reward-addr

0fea268

threads

e3fc776

threads 1

3a7aebd

keruch force-pushed the kirill/1248-sequencer-reward-addr branch from 5453a2f to 3a7aebd Compare October 16, 2024 13:54

keruch requested a review from mtsitrin October 16, 2024 13:54

keruch marked this pull request as draft October 16, 2024 18:23

keruch closed this Oct 18, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat(executor): added MsgUpsertSequencer consensus message #1120

feat(executor): added MsgUpsertSequencer consensus message #1120

keruch commented Oct 4, 2024 •

edited

Loading

faultytolly commented Oct 9, 2024

zale144 Oct 9, 2024

keruch Oct 10, 2024

keruch commented Oct 10, 2024 •

edited

Loading

keruch commented Oct 10, 2024

omritoptix left a comment

omritoptix Oct 12, 2024

omritoptix left a comment

keruch commented Oct 14, 2024 •

edited

Loading

mtsitrin Oct 15, 2024

mtsitrin Oct 15, 2024

mtsitrin Oct 15, 2024

mtsitrin Oct 15, 2024

mtsitrin Oct 15, 2024

keruch Oct 15, 2024

	sequencertypes "github.com/dymensionxyz/dymint/types/pb/rollapp/sequencers/types"
	rdktypes "github.com/dymensionxyz/dymint/types/pb/rollapp/sequencers/types"

feat(executor): added MsgUpsertSequencer consensus message #1120

feat(executor): added MsgUpsertSequencer consensus message #1120

Conversation

keruch commented Oct 4, 2024 • edited Loading

PR Standards

faultytolly commented Oct 9, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

keruch commented Oct 10, 2024 • edited Loading

keruch commented Oct 10, 2024

omritoptix left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

omritoptix left a comment

Choose a reason for hiding this comment

keruch commented Oct 14, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

keruch commented Oct 4, 2024 •

edited

Loading

keruch commented Oct 10, 2024 •

edited

Loading

keruch commented Oct 14, 2024 •

edited

Loading