Release Notes

0.3.1

Important

  • Bugfixes

    • Fix Canton topology import issue that can cause synchronizer upgrades with downtime to fail on some networks.

  • Deployment

    • Make the wallet sweep config value use-transfer-preapproval optional in the validator helm chart with a default of false.

0.3.0

Important

  • Daml recompilation may be required: this release changes the definition of the AmuletRules template arguments, as it introduces a new optional config value called transferPreapprovalFee (see Daml Changes in 0.3.0). If your Daml code depends on splice-amulet < 0.1.6, then you must recompile and redeploy it after the network was upgraded to splice-amulet-0.1.6 and before the SVs change this optional config value away from its default value.

  • This release must be applied through the Synchronizer Upgrade with Downtime procedure.

  • Canton

    This release upgrades from Canton 3.1 to Canton 3.2. In addition to stability improvements, the primary change is adding support for externally hosted parties, which enables supporting Amulet custody.

  • Validator App, Scan App

    • Add support for Amulet custody.

    • Fixed a bug where BFT scan calls would fail even though enough remote scan connections are available. This would happen if roughly a third of the SV nodes are offline.

  • Wallet UI

    • Support for non-external parties to exchange amulets with externally hosted ones via pre-approved transfers.

  • SV UI

    • The SV UI now shows a confirmation dialog when creating a Vote Request or Voting.

  • Deployment

    • An optional value uiPollInterval has been added to the Helm charts for splice-scan, splice-sv-node, and splice-validator. This value allows you to configure the interval at which the deployed UIs poll the services for updates in milliseconds. If unspecified, the default value is 1000 (1 second).

    • The log field labels."k8s-pod/cn-component" has been renamed to labels."k8s-pod/splice-component".

  • Security

    • Fix a Canton node initialization issue that caused newly initialized participants, mediators, and sequencers to reuse their root namespace signing key for all signing purposes. Upgrading to this release will also fix the key usage on all validators and SVs that were originally onboarded on an affected version (versions 0.2.4 to 0.2.8), generating fresh signing keys for affected Canton nodes.

  • Docs

    • Added a new section to the Validator documentation on how to share the operator wallet with multiple users. See Users, Parties and Wallets in the Canton Network Wallet.

    • Added a new subsection to Supervalidator documentation documenting the URL conventions agreed upon by the SV operators.

Daml Changes in 0.3.0

The Daml changes introduce support for the external custody of the keys of a Daml party. Signatures required from these external parties can be collected via a crypto custodian’s system, and can involve multiple human confirmers. Transactions submitted in the name of these parties can thus take multiple hours from the creation of the transaction signing request to the final commit of the transaction on the network. This increased latency required several changes in the Daml models underlying Amulet. They can be reviewed in detail by diffing the daml directory in the https://github.com/hyperledger-labs/splice repo.

The key changes are summarized below:

  • Changes the existing AmuletRules template:

    • Add a new config field transferPreapprovalFee in the AmuletConfig stored in AmuletRules.

      Important: once this field is set to Some value, you can no longer call choices on AmuletRules using Daml code built against a version before splice-amulet-0.1.6! Please recompile and redistribute your Daml code once the SVs have upgraded to splice-amulet-0.1.6 on your target network.

    • Add the choices AmuletRules_CreateTransferPreapproval and AmuletRules_CreateExternalPartySetupProposal explained below.

  • New workflows and templates:

    • Introduce the ability for a party to declare to the network that they are OK with receiving incoming Amulet transfers from any party by creating a TransferPreapproval. This is used by externally hosted parties to receive funds without having to actively confirm that they are OK to receive the funds. It must also be used by parties that want to receive funds from externally hosted parties, as external party wallets currently do not use the transfer offer workflow.

    • The TransferPreapproval contracts are expected to be created by the party’s crypto custodian, which pays the yearly maintenance fee. That fee is configurable via DSO vote and initially set to $1 per year. The payment itself happens by burning the corresponding amount of Amulet on purchase. In return for paying that fee, the crypto custodian is recorded as the app provider and validator operator on all Amulet transfers executed via the TransferPreapproval maintained by them.

    • A helper workflow called an ExternalPartySetupProposal has been added for crypto custody providers to set up both the TransferPreapproval and the ValidatorRight for an external party. The latter is required for claiming validator activity records. That workflow is initiated by the crypto custody provider calling the AmuletRules_CreateExternalPartySetupProposal choice.

    • Parties can also directly purchase a TransferPreapproval using AmuletRules_CreateTransferPreapproval choice.

    • Furthermore, parties are given the ability to delegate executing a Amulet transfer to a party of their choosing using the ExternalPartyAmuletRules_CreateTransferCommand. We introduced this feature because the normal Amulet transfer transactions refer to the OpenMiningRound contracts, which are valid for at most 30 minutes (10 minutes of pre-announcement time, and 2 * 10 minutes of active time). This time is too short to accommodate the human-in-the-loop confirmation workflows of crypto custody providers, which in turn would result in failed transactions due to referencing a stale round contract.

    • The typical choice for the delegate is a normal party on the crypto custodians node. That party is expected to be online and submit the actual transfer as soon as the TransferCommand is visible. The input amulets for the transfer are selected by the delegate; and they are expected to select inputs that cover the required amount provided they exist. In case there are not enough funds the TransferCommand gets archived and marked as failed.

    • External parties creating multiple TransferCommands are protected from executing the same transfer twice using an Ethereum style nonce tracked by the DSO, which must be sequentially increasing for a transfer command to be executed. We expect the wallet of these parties to select the right nonce using information available from Amulet scan. Having multiple transfer commands in-flight is supported.

    • All transactions involving TransferCommands and TransferPreapprovals have the dso party as a signatory and are thus always validated by ⅔ of the SV nodes.

  • The Daml changes in this release require a governance vote to upgrade the package configs to:

    name

    version

    amulet

    0.1.6

    amuletNameService

    0.1.6

    dsoGovernance

    0.1.9

    validatorLifecycle

    0.1.1

    wallet

    0.1.6

    walletPayments

    0.1.6

0.2.8

  • SV App

    • The query to fetch the vote results has been fixed for postgres 15.

  • Sequencer

    • Fix an inefficient query when querying the onboarding snapshot for a new SV that tries to onboard.

0.2.7

  • Scan

    • Added new endpoints /v1/updates and /v1/updates/{update_id}. The updates endpoint returns all Daml transactions and also all contract reassignments. Both Daml transactions and contract reassignments can be made up of multiple smaller components: A single Daml transaction may be the top node of a tree of sub-transactions, and a contract reassignment may actually be a batch of many reassignments.

      Each Super Validator node assigns a unique counter, called an event ID, to each of the sub-transactions in the Daml transaction tree. Because there’s not just one way to assign a counter to the elements of a tree, each Super Validator node gives different event IDs to the same elements of the transaction tree.

      This means that applications that want to compare updates from more than one Super Validator can’t match their event IDs. So for the v1 version of these endpoints, we’ve added a method for tree node numbering in Scan, which consistently produces the same event ids on each tree node, when given the same tree structure.

      Applications that rely on an existing set of event IDs drawn from a single Super Validator may continue to use /v0/updates and /v0/updates/{update_id}. This will return the single-Super Validator set of event IDs that they’ve used up to now. Applications that want to compare the details of updates, including transaction trees and sub-transactions, across Super Validators can use the v1 version of these endpoints.

0.2.6

Note: 0.2.5 was skipped as it introduced a regression where the splice apps hardcoded the wrong log level.

  • Docs

    • Updated docs to include a section on how to create a standalone k8s-based Canton Network. This can be useful to test deployment changes, in particular for SVs. See How to Bootstrap a Network.

  • SV UI

    • Configuration changes for AmuletRules and DsoRules are diffed against the configuration it will replace and the in-flights proposals. This makes it easier to see what changes are being proposed and what the current configuration is.

    • When creating validator onboarding secrets through the SV UI, they will now have an expiration time of 48 hours.

  • Scan

    • Added endpoint /v0/validators/validator-faucets to query the validator faucet by validator party Ids.

    • Modified the /v0/updates and /v0/updates/{update_id} Scan API endpoints to make sure they consistently returns the same history across SVs:

      • The /v0/updates endpoint now fails on scans that have not yet replicated history from before their SV node joined the network.

      • The /v0/updates endpoint now excludes updates resulting from ACS imports (those with workflow id starting with canton-network-acs-import).

      • Fix an issue where the ordering of stakeholders (signatories and observers) would be inconsistent across SVs when calling the /v0/updates and /v0/updates/{update_id} endpoints on the Scan API.

      • Fix a bug in /v0/domains/{domain_id}/members/{member_id}/traffic-status that resulted in the returned total purchased traffic value being incorrect after a hard migration.

  • Add a new index to Splice application databases. Scan and validator apps might take a while to start after the upgrade.

  • Canton

    • Enabled slow future logging for all components to better debug stuck nodes.

    • Added a max time of 10 minutes for processing of a sequenced event before the node crashes to get restarted. This mitigates cases where nodes might get stuck due to a bug and a restart recovers them.

  • Deployment

    • Breaking Every Helm chart with a name starting with cn- has been renamed, now starting with splice- instead, except for cn-docs.

    • Breaking The script token.py was renamed to get-token.py to avoid conflicting with some imported modules.

    • imagePullPolicy is now unset by default corresponding to IfNotPresent. You can overwrite it using the helm value imagePullPolicy if needed.

    • In paused-triggers settings, the trigger name prefix com.daml.network has been replaced by org.lfdecentralizedtrust.splice. This also applies to stacktraces you may see in logs.

    • domain.sequencerAddress, domain.mediatorAddress and participantAddress in the SV and Scan helm values are now mandatory. The defaults did not include the migration id so are almost always incorrect which means this likely has no impact as SVs should already have this set explicitly.

  • Bugfixes

    • Fix an issue in the wallet app where the transactions from previous migration ids would not be listed when paginating.

0.2.4

  • Sequencer

    Fix a rare bug where a lagging participant trying to submit a topology transaction resulted in the sequencer deadlocking and not processing any new events.

0.2.3

Note: 0.2.2 was skipped due to an error in the publishing process.

  • SV UI * The route to view the amulet price has been renamed from /cc-price to /amulet-price

  • The docker-compose validator now supports recovering from a node identities dump in case of a complete disaster.

  • Add new initialPackageConfigJson value to the SV helm chart to allow for setting the daml package version when bootstrapping a network. This is useful to ensure that the Daml versions do not change on a network reset. Only the first SV needs to set this.

  • SV app

    • Fix a bug where sequencer pruning treated nodes that have not joined after a synchronizer migration with downtime as lagging even when the pruning interval has not yet passed and disabled them preventing them from connecting to the sequencer.

  • Deployment

    • Breaking: The auth secrets splice-app-{sv,validator}-ledger-api-auth formerly had audience as an optional field. This is now required. The former implicit value was https://canton.network.global. If you have not overridden this value before, you should add it now explicitly.

    • It used to be possible to override the ledger-api audience value through the helm value auth.ledgerApiAudience in the sv and validator charts. This has been removed – use the secret mentioned in the previous point.

    • Breaking The chart value auth.audience was formerly optional, and is now required for the following charts. The previous implicit value was https://canton.network.global. To continue using it, please provide it explicitly to your values. (See the sv-helm and validator-helm docs for more information on auth configuration.) * cn-sv-node * cn-validator

    • Breaking The chart value auth.jwksUrl was formerly optional, and is now required for the same charts above. This should already be overridden in your values file for your particular auth setup, so likely no further action is required.

  • Bugfixes

    • Fix an issue where validators that were already deployed with an invalid validatorPartyHint were failing to start after a hard domain migration, as the already existing hint was rejected by the validator app.

  • Sequencer

    • Fix an issue in sequencer traffic management that resulted in a deadlock after a synchronizer upgrade with downtime where lagging validators failed to submit a transaction due to lagging behind but also failed to catch up due to the submission failing.

  • Added support for a docker-compose based deployment of a single-SV network, for app developers to test against without needing to connect to DevNet.

0.2.1

  • Added support for a docker-compose based validator deployment.

  • Scan

    • Fix an issue in the holdings and holding summary endpoint where it failed to decode contracts when the splice-amulet version the contract was created in did not match the latest supported version by the Scan release.

  • Sequencer

    • Fix a bug that prevented initialization during a hard domain migration if there was a proposal in the topology state on the old migration id.

0.2.0

Note: This release must be applied through the Synchronizer Upgrades with Downtime procedure.

  • Canton

    This release upgrades from Canton 3.0 to Canton 3.1. The primary change is a full redesign of the sequencer database to only store each sequenced messages once instead of duplicating it for each recipient.

  • Daml

    • Add a choice that allows merging duplicated validator licenses. On DevNet it is easy to get duplicates as secrets can be automatically generated

    • by querying the /api/sv/v0/devnet/onboard/validator/prepare endpoint. This is not an issue on Test/MainNet where secrets are explicitly provisioned by SV operators and are one-time use.

    • It is up to the SV operators to ensure that they only hand out one secret to each validator

    • Add a new template ValidatorLivenessActivityRecord. It is a copy of the ValidatorFaucetCoupon template with the only difference being that the validator is an observer instead of signatory. This is to allow to expire the coupon without the validator’s involvement.

    • The Daml changes in this release require a governance vote to upgrade the package configs to:

      name

      version

      amulet

      0.1.5

      amuletNameService

      0.1.5

      dsoGovernance

      0.1.8

      validatorLifecycle

      0.1.1

      wallet

      0.1.5

      walletPayments

      0.1.5

  • SV and validator apps

    • Add a note about avoiding installing third-party Daml apps on SV nodes in the SV operations documentation, as that may compromise the security of the SV node.

    • Remove support for deprecated bootstrapTXs field on node identity dumps. Node identity dumps taken on a 0.1.2 snapshot or earlier version are no longer supported.

  • Metrics: All the histograms default to using native histograms.

    • Dashboards were also adjusted to use the PromQL functions for native histograms in all the queries

    • You can turn off this behavior for each component by adding the following env variable in the additionalEnvVars helm values: ADDITIONAL_CONFIG_DISABLE_NATIVE_HISTOGRAMS=”canton.monitoring.metrics.histograms=[]”

  • Dashboards

    • Added a new “Synchronizer Fees (Validator View)” dashboard for validators to monitor their traffic purchases and consumption.

  • Wallet API

    • The list API in wallet-internal.yaml now exposes contracts as ContractWithState instead of just as a Contract.

  • Deployment

    • Removed the disableAutoInit value from the helm charts of Canton nodes. All Canton nodes will now always start with initialization disabled. SV and validator apps will take care of initializing the nodes as needed, using use the new nodeIdentifier helm chart value for the Canton node identifiers. The installing instructions for validators and SVs have been updated accordingly.

    • spliceInstanceNames values are now mandatory for all Helm charts that deploy a frontend (cn-scan, cn-validator, cn-sv-node, and cn-splitwell-web-ui). The correct values for them are published in the docs for validators and SVs.

    • The configuration variable clusterUrl was removed from all Helm charts except splitwell-web-ui.

    • Default Postgres PVC size for validators is configured as 50GiB in the new postgres-values-validator-participant.yaml examples file. Note also the change in the validator installation docs to use this file while installing the Postgres chart.

    • For the Docker images, these input environment variables have been renamed, replacing CN with SPLICE:

      • CN_APP_UI_HTTP_URL

      • CN_APP_UI_UNSAFE_SECRET

      • CN_APP_UI_UNSAFE

      • CN_APP_WALLET_REDIRECT

    • The Kubernetes secrets below have been renamed, replacing cn- with splice-:

      • cn-app-*-ledger-api-auth

      • cn-app-cns-ui-auth

      • cn-app-sv-key

      • cn-app-sv-ui-auth

      • cn-app-validator-onboarding-validator

      • cn-app-wallet-ui-auth

  • Documentation

    • Updated recommendations for checking synchronizer health after a Synchronizer Upgrade with Downtime to focus exclusively on monitoring signals.

    • Simplified jq-based data dump post-processing examples in disaster recovery documentation for SVs and validators.

  • Metrics

    • Added cn_wallet_unlocked_amulet_balance and cn_wallet_locked_amulet_balance metrics to expose the effective per party balance of locked and unlocked amulets.

0.1.19

  • Fix the Docker image digest which was used for the ans-web-ui and accidentally was empty (thereby not pinning the image) in 0.1.18 due to a rename.

  • validatorPartyHint is now mandatory for non-SV validators. For an existing validator, it must be set to the current party hint (otherwise, the app will fail to start). For new validators, it must be of format <organization>-<function>-<enumerator>, where organization and function are alphanumeric, and enumerator is a number starting from 1.

  • Fix an issue in the scan ACS snapshot functionality added in 0.1.18 for network bootstrapped just before 0:00.

  • Fix an issue in the ACS snapshot functionality added in 0.1.18 around hard domain migrations. This only affects a hard domain migration to 0.1.18 but not from 0.1.18.

0.1.18

  • SV apps

    • Fix a rare race condition where the SV app uses the wrong timestamp to export the topology state on a hard domain migration resulting in the sequencer failing to initialize after the migration. We recommend upgrading before the next hard domain migration.

    • Enable SV to retain pre-migration sequencer URLs in SvNodeState. This is done through a new migration.legacyId configuration in the SV values. If set, the SV will keep exposing its sequencer URL for that migration id. Once you undeploy the old sequencer node, remove this option as well to stop Scan from advertising your old sequencer. This allows validators that have been lagging behind to catchup easier.

  • Dashboards

    • Added a new CometBFT Network Status dashboard that displays how much data is being exchanged with each peer on the CometBFT P2P network. This should should make it easier to diagnose connectivity problems between network peers.

  • Scan API

    • Added the getUpdateById API in scan-internal.yaml. The getUpdateById API can be used to retrieve an update by its update ID.

    • Added the getAcsSnapshotAt, getHoldingsStateAt and getHoldingsSummaryAt APIs in scan-internal.yaml. A snapshot of the active contract set (ACS) is now computed and stored periodically to serve these endpoints.

    • Modified listDsoSequencers Scan API to also expose pre migration sequencer urls, allowing pre-migration validators to catch up.

  • UI

    • Gzip compression has been enabled for the Scan, Wallet, SV and CNS UIs.

  • Deployment

    • Updated the Cometbft Helm chart to not accept integer values for the chainIdSuffix.

    • The disableAutoInit Helm value now defaults to true wherever it is used and must be explicitly set to false when onboarding fresh validators or SVs. The installing instructions for validators and SVs have been updated accordingly.

    • Added helm.sh/resource-policy: keep to validator and SV app domain migration PVCs to ensure they don’t accidentally get deleted by a helm uninstall. You can still fully delete them with a kubectl delete pvc.

    • validatorPartyHint is now mandatory for non-SV validators. For an existing validator, it should be set to the current party hint (otherwise, the value will be ignored, and a warning will be printed to log). For new validators, it should be of format <organization>-<function>-<enumerator>.

    • In cometbft-values.yaml, the top-level label founder is now sv1. The example has been updated to match, and this change must be made to your own copy.

    • The download link for the release bundle has changed to a new URL format: <version>_splice-node.tar.gz. Its content has been renamed accordingly as well.

  • Documentation

    • Simplified jq-based data dump post-processing examples in disaster recovery documentation for SVs and validators.

0.1.17

  • Wallet automation

    • Fix an issue in the wallet sweep automation where it created additional transfer offers even if there were already sufficient transfer offers to cover the sweep.

  • Deployment

    • Image versions in Helm charts are now pinned to digests for extra security

0.1.16

  • CometBft

    • The default cometbft persistent volume size was bumped from 1250Gi to 2500Gi.

  • SV app

    • Add automation to automatically call the Daml choice that prunes futureValue added in 0.1.15

  • Release

    • HTML docs are now included in the release bundle, under docs/html.

  • Documentation

    • Added notes about configuring traffic top-ups for validators to validator-values.yaml

  • Daml

    • Fixed a bug in AmuletRules_ComputeFees where the fee computation for locks was too high as it did not do the same deduplication of lock-holders as is done by AmuletRules_Transfer.

    • Fixed ANS entry expiration so that it’s robust to stakeholder participants being unavailable.

    • All Dars have been rebuilt from source files that include the same copyright prefix as in the Splice repository. This bumps dar versions in all packages. Incorporating that will require a governance vote to upgrade the package configs to:

      name

      version

      amulet

      0.1.4

      amuletNameService

      0.1.4

      dsoGovernance

      0.1.6

      validatorLifecycle

      0.1.1

      wallet

      0.1.4

      walletPayments

      0.1.4

  • Deployment

    • Added an livenessProbeInitialDelaySeconds parameter to all helm charts.

    • Helm charts that deploy a frontend (cn-scan, cn-validator, cn-sv-node, and cn-splitwell-web-ui) now accept a new parameter, spliceInstanceNames, to configure network-specific terminology. The correct values should be consumed from the cn-svc-configs ui-config-values.yaml

    • Docker environment variables of the form CN_APP_*_UI_* have been renamed to CN_APP_UI_*, dropping the app name prefix. For users of the Helm charts, no further action is needed.

  • Sequencer

    • Improve performance of sequencer startup and querying the sequencer onboarding snapshot when onboarding new SVs. This adds a new index to the sequencer database so can take a while depending on the size of the DB.

      Note: If you encounter issues with the migration taking too long and k8s killing your pod, bump the livenessProbeInitialDelaySeconds parameter in the sequencer helm chart.

      We have also seen some issues with istio cancelling the database connection before the migration can finish (on much larger scale clusters than what we expect to have on dev/test/mainnet). In that case, consider disabling the istio proxy through annotations: traffic.sidecar.istio.io/excludeOutboundPorts: "YOURDATABASEPORT" on the sequencer deployment.

  • All helm charts now allow configuring the database port through persistence.port. Note that for the cn-global-domain chart, this is nested under sequencer.persistence and mediator.persistence.

0.1.15

Note: 0.1.14 was skipped as it contained an issue related to logging. Upgrade directly from 0.1.13 to 0.1.15.

  • SV app

    • Added a governance option to update the SV reward weight of a member SV. This is available in the Governance tab by selecting the action “Update SV Reward Weight”.

    • Added consensus_state to the list of CometBFT RPC endpoints exposed via the SV app at /v0/admin/domain/cometbft/json-rpc.

  • Deployment

    • Fix an issue in the validator and SV helm charts where setting contactPoint to an empty string produced an error.

  • Daml

    • Add a choice that allows pruning configs from the AmuletRules futureValues after the time has been reached to reduce the size of the config and reduce differences between the config schedule on different networks.

    • The Daml changes in this release require a governance vote to upgrade the package configs to:

      name

      version

      amulet

      0.1.3

      amuletNameService

      0.1.3

      dsoGovernance

      0.1.5

      validatorLifecycle

      0.1.0

      wallet

      0.1.3

      walletPayments

      0.1.3

0.1.13

  • Docker

    • Switch to using eclipse-temurin:17-jdk-jammy as the base image as the openjdk:17-jdk-slim is no longer maintained.

  • Deployment

    • UI containers in the Helm charts now request only 0.1 CPU and 240Mi memory by default.

    • Default participant CPU requests have been lowered from 2 to 1 CPU based on the observed usage under load tests.

    • Validator and SV helm charts have a new required contactPoint field that must be set in validator-values.yaml and sv-values.yaml. This should point to a Slack username or email address that can be used by other node operators to contact you in case there are issues with your node. If you do not want to share this, set it to an empty string.

    • Added support for k8s tolerations to all Helm charts.

  • SV app

    • /v0/admin/domain/data-snapshot now includes created_at and migration_id in the response payload, so these no longer need to be added manually when restoring an SV app from backup. migration_id is also an optional argument to set the latter, defaulting to 1 + the cluster’s current migration ID.

    • The extra beneficiary config has been changed to specify weights in an ordered list instead of percentages. The weights are distributed in the order of the list until there is no weight remaining. Any remainder still goes to the SV operator party. This fixes two problems with the percentage-based beneficiary specification:

      1. it does not suffer from rounding errors

      2. it allows changing the config ahead of time to account for a planned weight changes by adding additional entries at the end.

      This is a breaking config change, which requires you to adapt the SV app config as per this example: assuming a total weight of 10000 basis points, the previous config:

      extraBeneficiaries:
  - partyId: "BENEFICIARY_1_PARTY_ID"
    percentage: 10.0
  - partyId: "BENEFICIARY_2_PARTY_ID"
    percentage: 33.33

      changes to:

      extraBeneficiaries:
  - beneficiary: "BENEFICIARY_1_PARTY_ID"
    weight: 1000
  - beneficiary: "BENEFICIARY_2_PARTY_ID"
    weight: 3333

  • Validator app

    • /v0/admin/domain/data-snapshot now accepts migration_id as an argument, overriding migrationId in the response payload. The default migrationId is now 1 + the cluster’s current migration ID, rather than only the current migration ID.

    • The migration dump format has changed; the JSON keys acsSnapshot, acsTimestamp, migrationId, domainId, and createdAt have changed to acs_snapshot, acs_timestamp, migration_id, domain_id, and created_at, respectively. The format of /v0/admin/domain/data-snapshot has been fixed where it mismatched the migration dump import format so that backups do not need to be patched to be restored. Previous dumps can still be imported using the old format.

  • Scan app

    • Improved performance of the per-party ACS endpoint that is used when reonboarding a validator from the identity backup.

  • Daml

    • Extended the Daml models to report the version number and a periodic heartbeat of each validator to provide a better overview of the network state and detect potential issues from upgrades earlier.

    • The frequency of ACS commitments can now be modified via a “Set DsoRules configuration” governance by changing the newly added acsCommitmentReconciliationInterval configuration parameter in the DsoRules (set by default to 30 minutes).

    • Removed a special case for SRARC_OffboardSv in the DsoRules_CloseVoteRequest choice in splice-dso-governance.dar, so that offboarding an SV before the vote request expires is now only possible if all current SVs agree, including the SV that is being offboarded. Prior to this change, the offboarding would become effective before the set expiration time once all SVs except the SV to be offboarded had voted. This complicated the coordination around giving SVs sufficient time to address the offboarding reason and prevent the offboarding.

    • The Daml changes in this release require a governance vote to upgrade the package configs to:

      name

      version

      amulet

      0.1.3

      amuletNameService

      0.1.3

      dsoGovernance

      0.1.4

      validatorLifecycle

      0.1.0

      wallet

      0.1.3

      walletPayments

      0.1.3

  • Dashboards

    Added a new Validator License dashboard that displays the version and contact point of all validators. This can be useful to judge the impact of an upgrade.

0.1.12

Note: 0.1.11 was skipped as it contained some issues. Upgrade directly from 0.1.10 to 0.1.12.

  • SV and Validator app

    • Added a disableIngestUpdateHistoryFromParticipantBegin flag to the helm values of the SV and validator app. This was added to account for a change in 0.1.11 that stores more history as backfilling the history on the existing test/devnet clusters is too expensive. This should only be enabled on existing Dev/TestNet clusters to avoid issues when upgrading to 0.1.12. It must not be enabled on any new cluster or if a node is fully reset.

  • Scan

    • Fix a bug where the new update history API in scan was unable to serve data from before the upgrade.

  • Include Grafana dashboards and a README on network health in the release bundle.

  • Configuration

    • Add support in the Validator app Helm chart for configuring sweeps and auto-accepts of transfer offers.

    • The wallet-sweep and auto-accept configuration values for a validator app were changed to map party-ids to configurations instead of mapping participant user-names to configurations.

  • Daml

    • The WalletAppInstall_ExecuteBatch choice in splice-wallet.dar was changed to also record the wallet user party when executing batches of operations on a user’s coin holdings to improve disambuiguation of log entries in the wallet transaction log.

    • Fix an issue in the computation of transfer fees where the values of the steps were interpreted as the difference between steps as opposed to an absolute value so e.g. the fees were computed as transferFee(2000) = 0.1 * 100 + 1000 * 0.01 + 900 * 0.001 instead of transferFee(2000) = 0.1 * 100 + 900 * 0.01 + 1000 * 0.001 for the default config.

      This requires a governance vote to upgrade the package configs to:

      name

      version

      amulet

      0.1.2

      amuletNameService

      0.1.2

      dsoGovernance

      0.1.3

      validatorLifecycle

      0.1.0

      wallet

      0.1.2

      walletPayments

      0.1.2

  • Validator admin API

    Simplified creating users that share the same party-id and wallet. For that purpose POST /v0/admin/users accepts an optional party_id field in its JSON body, which can be set to an already allocated party.

  • Bugfixes

    • The wallet automation for collecting rewards is started only once per Daml party instead of once per onboarded wallet user. This enables setups where multiple wallet users have access to the same coin holdings for the same Daml party.

    • Fixed a bug where a user wallet wrongly attempted to use the featured app right of the validator admin party if that existed, which resulted in failed transactions.

  • The approved-sv-id-values-*.yaml files have been removed from the release bundle. The approved SV identities for each network instance can now exclusively be obtained from the cn-svc-configs repo .

  • CC Scan

    Fix a bug in the balance API and UI where balances did not get tracked properly if the balance change for a given party was negative in one round, e.g., because it transferred away a large amount.

0.1.10

  • SV App

    The default transfer config set by the founding node has been changed from "0.0000192901 to 0.0000190259 corresponding to changing the computation to be performaned in fixed point decimals and 365 days. This matches the change already applied to devnet through a governance vote.

  • Daml

    Fixed a bug that resulted in duplicate SvRewardState contracts when an SV got reonboarded which allowed them to receive rewards corresponding to a multiple of their actual weight. This requires upgrading dso-governance to 0.1.2 through a governance vote on AmuletConfig.

  • SV UI

    Fixed a bug in pretty printing of the JSON object in DSO Info that printed maps differently from the API response and some other parts of the UI.

0.1.9

  • Configuration

    • Default actionConfirmationTimeout parameter in CoinRules was increased from 5 minutes to 1 hour. This increases robustness if some nodes are temporarily unavailable or slow. Note that this requires a governance vote to change the DsoConfig on existing clusters.

    • Default PVC sizes updated: 2800Gi for Postgres.

  • App Dev

    • DARs can no longer be uploaded through the Ledger API. Instead use the Canton admin API. This change was made as the ledger API upload breaks under hard domain migrations.

  • Documentation

    • Add notes about (Helm chart) version upgrades to the Synchronizer Upgrades with Downtime documentation sections for SVs and validators.

    • Updated Preparing for Validator Onboarding sections to describe the steps a validator operator needs to take to onboard a new node.

    • Removed Self-Hosted Validator documentation in favor of the Helm docs for validator deployments.

    • Removed Splitwell-related documentation as Splitwell is not actively maintained as a production-ready app.

  • Deployment

    • The values nodeId, publicKey and keyAddress in the founder section of the cometbft helm chart are not set in the chart defaults but must be explicitly provided. See the comments in the example cometbft-values.yaml for the values to use for DevNet, TestNet or MainNet.

  • Daml

    Fixed a bug that prevented a round from moving to the issuing state if there are no unclaimed rewards for that round. This requires upgrading splice-amulet, splice-amulet-name-service, splice-dso-governance and splice-wallet to version 0.1.1 through a governance vote on AmuletConfig.

0.1.8

  • Deployment

    • The URL for the Digital-Asset-2 node is now compliant with the agreed upon URL formats: *.sv-2.<dev|test>.global.canton.network.digitalasset.com

    • All Digital-Asset-Eng-X nodes also change URLs with this release, from *.sv-x.<hostname> to *.sv-x-eng.<dev|test>.global.canton.network.digitalasset.com.

  • Bugfixes

    • Reduced the frequency of ACS commitments to every 30min to avoid issues with validators running out of traffic.

  • Performance

    • Sequencers now batch some of their writes which should improve performance.

0.1.7

  • Deployment

    • Note change in urls in the Digital-Asset-2 node (which is used in several example and default configurations in the docs), from *.sv-1.svc.<hostname> to *.sv-1.<hostname>, as a step towards making that node compliant with the agreed upon URL formats. Note that further changes to Digital Asset node URLs might become effective before the next release becomes available.

  • Updated validator runbooks with instructions for re-onboarding a validator.

  • Renamed traffic-reserved-for-topups in the validator app and SV app config to reserved-traffic to better reflect the fact that the “reserved” traffic amount is used for more than just traffic top-ups. No change is needed unless you explicitly set a value for this instead of just relying on the default.

  • APIs

    • The admin/domain/data-snapshot endpoints on the SV and validator app now require specifying the timestamp as a query parameter instead of in the payload body. This was changed since GET requests must not have request bodies.

0.1.6

Note: 0.1.5 resulted in the issue mentioned below so both SVs and validators should directly upgrade from 0.1.4 to 0.1.6.

  • Security

    • Fixed an issue where secrets in config files were logged on startup. This effects Auth0 secrets, SV onboarding and validator onboarding secrets. Please rotate all those secrets as soon as possible to reduce the impact.

  • Bugfixes

    • Fix a bug (triggered by some changes in 0.1.5) where automation could submit too many commands in parallel overloading the synchronizer.

0.1.5

  • Fixed the SV UI to show node status information in the DSO info tab and display AmuletConfigChange vote requests that were executed.

  • Removed PVC size overrides in example postgres-values-participant.yaml and postgres-values-sequencer.yaml files. The Postgres instances used by the participant and sequencer should use the default size instead (1300Gi).

  • Updated the scan UI to show recent activity in a way that is more consistent and matches the actual activity on the ledger. Note that all transfers recorded in the past will show as having no sv rewards. This limitation can be removed with a future update.

  • Fix a bug where the namespace triggers did not get started on SV’s with migrating: true which prevented new SVs from being onboarded after domain migrations.

  • Updated SV and validator runbooks with network-wide disaster recovery instructions.

  • Introduced a vpns section in the IP whitelists json file, replacing the infra.vpn one.

0.1.4

  • Default PVC sizes updated: 640Gi for CometBFT and 1300Gi for Postgres.

  • Bugfix in Total balance and Total rewards in USD in Scan UI.

  • New value for cometbft-values.yaml: genesis.chainIdSuffix. Please explicitly set this to "0" as per the updated example. Note that this deprecates genesis.chainIdVersion, which can be removed for deployments that use this and later releases.

  • By default, CometBFT deployments now use the premium-rwo storage class for increased performance. Please override db.volumeStorageClass in your cometbft-values.yaml if this storage class is not supported by your Kubernetes cluster provider. Please use an SSD storage class for the CometBFT PVC.

  • Updated SV runbook for Re-onboarding an SV.

0.1.3

  • The Scan frontend shows information about currently open mining rounds in the current configuration box.

  • Minor documentation improvements related to synchronizer upgrades with downtime.

  • Fixed the initial validator rewards tranche to be 5% of the total issuance (it was wrongly set to 50%). Note that this only has an effect on newly bootstrapped clusters. Existing clusters need to be changed through a voting process.

  • Set the validatorFaucetCap explicitly to 2.85 instead of leaving it unset to make reviewing the config easier. This has no effect since unset defaults to 2.85. Existing clusters need to be changed through a voting process.

  • The resource requests for sequencers have been increased to match our target scale. If needed, they can be reduced using the sequencer.resources value of the cn-global-domain but please try to get them to a comparable value in time for mainnet.

  • Fix a sequencer bug that resulted in it failing to process any further messages after a message with high traffic costs.

  • If a tap fails in the wallet frontend, the error message includes extra technical details that may be useful for diagnosis.

0.1.2

  • Fixed a bug where coins with very large values broke ingestion in the SV and validator app due to an overflow.

  • Updated SV runbook for correct recommendation on pruning intervals.

2024-04-01

  • Renamed the following terms in our underlying Daml models and the apps’ APIs to prepare for open-sourcing their code in a form that does not use the term “Canton” or “collective”:

    • Coin -> Amulet

    • CNS -> ANS (Amulet Name Service)

    • SVC -> DSO (Decentralized Synchronizer Operations)

    • Domain -> Synchronizer

    • Global Domain (whenever it refers to the more generic concept) -> Decentralized Synchronizer

    • Note that for technical reasons the URLs for networks still include the term “svc” for now; e.g., https://wallet.sv.svc.YOUR_HOSTNAME.

  • Added an option to disable the Validator apps’ wallet. This can be done by setting enableWallet to false in the validator-values.yaml file.

  • Added ANS name resolution (formally known as CNS) for dso.ans to the DSO party and <sv-name>.sv.ans to all SV members parties.

  • CometBFT pruning duration has been increased to 30 days. No configuration changes are required.

  • Sequencer pruning period has been adjusted to 30 days and pruning interval has been reduced to 1 hour. Adjust sequencerPruningConfig.pruningInterval and sequencerPruningConfig.retentionPeriod in your sv-values.yaml to match the example sv-values.yaml.

  • The sequencer URL of the Digital Asset 2 node https://sequencer.sv-1.svc.CLUSTER.network.canton.global is no longer exposed. Instead use https://sequencer-MIGRATION_ID.sv-1.svc.CLUSTER.network.canton.global where MIGRATION_ID is the current migration id of the cluster.

  • Round 0 now has a duration of 26h. The two extra hours are to allow for internal validation before the release is announced while still providing 24h for anyone else to validate the config.

  • DevNet and TestNet are deployed on Monday instead of Sunday each week.

2024-03-25

  • Round 0 now has a duration of 24h. This removes the advantage of early joiners and allows for more time to validate that the configuration upon joining is the one an SV expected.

  • Initial coin price is now $0.005/CC.

  • Subsequent round duration is now 10min.

  • The initial holding fee is now $0.0000192901/round (about 4× its prior value) to preserve an approximate fee of $1/360 days given the round duration change.

  • New initial-holding-fee setting for "found-collective" sv onboarding.

  • Sequencer pruning is now enabled by default. This requires configuring a pruning interval and retention period in the SV app’s configuration.

  • Fix performance bottleneck while initializing new synchronizer after a hard synchronizer migration.

  • Fix scan so that it functions as expected after a hard synchronizer migration.

2024-03-18

  • Deployment

    • participant-values.yaml and global-domain-values.yaml now require specifying your SV name as nodeIdentifier: YOUR_SV_NAME. This is used to provide better names to Canton nodes.

    • Multiple changes to the way (non-SV) validator nodes are deployed, to prepare for supporting Synchronizer Upgrades with Downtime. Please revisit the section on Helm-based validator deployment, paying attention to the new MIGRATION_ID variable (should be set to 0 until further notice).

  • Documentation

  • Configuration

    • SV node renames:

      • Digital-Asset is preparing to run two nodes, Digital-Asset-1 and Digital-Asset-2

      • Digital Asset engineering team’s extra nodes on DevNet were renamed to Digital-Asset-Eng-X

    • SV weights: The SV weights on DevNet have been updated

2024-03-11

  • Deployment

    • Multiple changes to the way SV nodes are deployed, to prepare for supporting Synchronizer Upgrades with Downtime. Please revisit the section on Helm-based SV deployment, paying attention to the new MIGRATION_ID variable (should be set to 0 until further notice).

    • sv-values.yaml now also requires you to specify an internalUrl for your scan instance that the SV app can use to query its status.

    • In preparation for the mainnet deployment and testing real upgrades, testnet no longer preserves coin balances and validator licenses. No configuration changes are required for this. However, any validator secrets created through the UI or API now need to be regenerated on each reset. Validator secrets configured in expected-validator-onboardings will automatically be recreated. Note that this affects only testnet so will only take effect on March 18th.

  • Documentation

  • The SV rewards are now issued in accordance to CIP-0001.

2024-03-04

  • Deployment

    • It is no longer necessary to specify anything related to globalDomain in participant-values.yaml.

  • Documentation

    • Added section on Synchronizer Upgrades with Downtime. This section only contains a high-level overview for now and will be expanded in the upcoming weeks.

    • Preliminary documentation of restoring from backups. Note that for now, only the case of restoring a full SV node from a backup is fully covered.

  • SVs do not pay domain fees anymore for their nodes. Therefore, traffic top-ups do not need to be configured for SV validators. In keeping with this, topup.enabled in sv-validator-values.yaml is set to false.

  • The RemoveMember action has been renamed to OffboardMember in the SV UI and SvcRules Daml model, including SRARC_OffboardMember and SvcRules_OffboardMember.

2024-02-26

  • Deployment

    • Removed option to configure Kubernetes node affinity for PVCs due to a faulty implementation. For controlling the provisioning of PVCs, you can define custom storage classes and configure them via the respective db.volumeStorageClass Helm chart field.

    • Fix the affinity and nodeSelector field on the cn-postgres Helm chart so they are applied as expected.

    • The scanAddress in validator-values.yaml should be an address to a trusted Scan instance that is reachable by your Validator.

  • Documentation

    • Added instructions for fetching and backing up node identities from an SV node in the Backups section.

2024-02-19

  • The scan app is now initialized with last computed aggregates from other scans in the SVC.

  • Deployment

    • You can now configure Kubernetes affinity and node selection rules for pods deployed as part of CN helm charts. This is done by setting the affinity and nodeSelector fields in the Helm values files, respectively. For helm charts that deploy persistent volumes, you can additionally configure Kubernetes node affinity for those volumes. This is done by setting the db.volumeNodeAffinity field in the respective Helm values files. For all of these fields, the standard Kubernetes configuration syntax applies. See also the examples given (as commented-out lines) in cometbft-values.yaml.

2024-02-12

  • The JSON encoding of jsonb columns in the database has been changed. Please make sure to clean all the database before upgrading. Data with the old JSON encoding cannot be read by the new version of the software.

2024-02-05

  • The wallet of non-SV validators now execute all reads through a Scan proxy in the validator, thus executing them in a BFT fashion.

  • You might see some ACS_COMMITMENT_MISMATCH warning logs in the participant. These can be ignored.

  • Add enableHealthProbes to the global domain helm chart, providing the ability to disable gRPC readiness and liveness probes for the sequencer and mediator.

  • Containers now use tini as the entrypoint to ensure proper signal handling.

  • Fix for wallet balances incorrectly reporting as zero for rounds that have not been aggregated yet. An error will be returned instead.

  • Deployment

    • The postgres instance has been split into four different instances: sequencer-pg, mediator-pg, participant-pg, apps-pg. Please see the new section on installing and configuring Postgres: Installing Postgres instances

    • postgres PVC size are new set to 480GB for the sequencer, and 48GB for each of the other three instances.

    • The default database names for the different components have been changed to cantonnet_<componentName>. They are all created automatically in init containers attached to the respective app pods.

2024-01-29

  • The /v0/wallet-balance endpoint to query a party’s CC balance is exposed through the Scan app.

  • Non-SV validators now connect to all registered Scans in the domain and read from them in a BFT fashion (safe as long as less than 1/3 of them are faulty). Scans are registered as part of the SV onboarding, using the public URL that was configured in the 2024-01-15 deployment (scan.publicUrl in SV-app’s helm chart). SV validators still trust the Scan app of the SV.

  • Clean up our OpenAPI spec to use oneOf. This involves minor changes to the some of our API endpoints including:

    • getHealthStatus for all nodes

    • getBuyTrafficRequestStatus and getTransferOfferStatus in the wallet API

    • getCometBftNodeStatus and getCometBftNodeDebugDump in the SV app

    In almost all cases, the changes should only involve some fields that were optional now becoming required.

2024-01-22

  • Adjust traffic purchase rewards structure to CIP-2:

    • Validator rewards are issued over the full amount of CC spent.

    • No app rewards are issued for traffic purchases.

    • Set the minimum traffic purchase amount to 1 USD to ensure coverage of execution cost.

    • Issue an extra app reward over $1 for CC transfers facilitated by featured apps.

  • Deployment:

    • Enabled pruning in sequencers of canton foundation

2024-01-15

  • Deployment:

    • The SV-app helm chart now expects scan.publicUrl to be set to the URL of the Scan app.

    • The syntax of the Helm charts configuring persistence has been standardized through the different images. The provided snippet highlights the optional fields and its syntax. To find the expected values for each image, refer to the _required.yaml_ Helm files.:

      sequencer:
  driver:
    type:
    port:
    host:
  persistence:
    databaseName:
    host:
    port:
    user:
    secretName:
mediator:
  persistence:
    databaseName:
    host:
    port:
    user:
    secretName:

  • Documentation:

    • Removed the section Renaming an SV in the page “Kubernetes-Based Deployment of a Super Validator node”. Not longer possible through the JSON API.

2024-01-08

  • Deployment:

    • The global domain helm chart now supports separate Postgres instances for the sequencer and mediator. They can be configured in the sequencerPostgres and mediatorPostgres values. The single postgres value has been deprecated and is no longer supported. (It is still possible to use a shared postgres instance, by configuring it under both sequencerPostgres and mediatorPostgres.)

    • The CometBFT egress ports have changed from 26656, 26666, 26676, 26686, 26696 to 26016, 26026, 26036, 26046, 26096

    • The CometBFT founder port was updated in the cometbft-values.yaml file to 26016 (from 26656). This is found under the founder.externalAddress field.

  • Bugfixes:

    • CometBFT state sync has been fixed after the recent issues and can once again be used for fast CometBFT node bootstrapping. (It should not be necessary anymore to override stateSync.enable to false in cometbft-values.yaml.)

2023-12-18

  • Minor SV UI tweaks. Among other things, it is now mandatory to supply a textual proposal summary when submitting a vote request.

  • Renamed directory to CNS across the system

    • Renamed ingress rule from https://directory.sv.svc.<YOUR_HOSTNAME>* to https://cns.sv.svc.<YOUR_HOSTNAME>*. Please note that this is now a requirement for this UI to continue working properly.

2023-12-11

  • Deployment:

    • The https://directory.sv.svc.<YOUR_HOSTNAME>/api/json-api/* ingress rule is no longer required for validators and super-validators

    • The helm charts now allow configuring the secret in which the postgres password is stored. Default is postgres-secrets.

  • Documentation:

    • Added an ingress rule for https://directory.sv.svc.<YOUR_HOSTNAME>/api/validator, which was accidentally omitted from the instructions.

2023-12-04

  • The party representing the supervalidator collective is renamed from svc::... to SVC::..., for consistency with SV party names.

  • The password for the PostgreSQL database is now set in Kubernetes secrets as opposed to Helm values files. Please refer to the updated documentation on Configuring PostgreSQL authentication for SV operators and Validator operators.

  • Documentation:

    • Clarify that using custom auth audiences is recommended.

    • Updated SV runbook to explain also ip white-listing.

    • Updated SV runbook with a list of outbound traffic.

  • The /v0/activities and /v0/transactions Scan APIs now include normalized balance changes per party. Please see the Scan OpenAPI specification.

2023-11-27

  • The domain.sequencerPublicUrl configuration in cn-sv-node helm chart is now mandatory. All SVs must specify the URL at which their sequencer can be reached.

  • The Canton Name Service (CNS) application is now decentralized. Rather than the founder SV operating a backend service for CNS, the application is now a decentralized one, operated by the SVC with BFT guarantees. No deployment or configuration changes are required for this change.

  • CNS entry name limit is changed from 40 to 60 (including suffix).

  • The SV, Scan, directory and validator apps now require a PostgreSQL database to run. This can be configured in sv-values.yaml, validator-values.yaml, scan-values.yaml.

2023-11-20

  • The CometBft RPC endpoints required for state sync are now exposed through the SV App. Accordingly, the list of rpcServers in cometbft-values.yaml has been updated to reflect the new URL path.

  • Documentation fixes

2023-11-13

  • CometBft pruning is now enabled, with approximately 7 days of history retained by default.

  • Revised the SVC governance formula, so that we now need ceil( (n + f + 1) / 2.0 ) SVs to form a quorum, where f := floor ( (n - 1) / 3.0 ) is the maximum supported number of faulty SVs for safe operation.

  • Removing the adjustment of SVC governance thresholds for DevNet, aligning it with the one used in TestNet.

  • Deployment updates:

    • The URL of the global domain sequencer hosted by the Canton Foundation has changed to https://sequencer.sv-1.svc.<TARGET_CLUSTER>.network.canton.global. This change is reflected in the values specified in participant-values.yaml, validator-values.yaml and sv-values.yaml.

    • The requirement for URL rewriting in the rules for Scan and SV apps has been removed: https://scan.sv.svc.<YOUR_HOSTNAME>/api/scan and https://sv.sv.svc.<YOUR_HOSTNAME>/api/sv no longer requires rewriting (and also has been modified from /api/v0/scan to /api/scan and from /api/v0/sv to /api/sv). For example, https://scan.sv.svc.<YOUR_HOSTNAME>/api/scan/foobar should be forwarded to http://validator-app:5012/api/scan/foobar. Note that URL rewriting is now required only in the ingress rule of the JSON API used by the directory frontend. This rule will be completely removed in the future.

    • Note that the readiness and liveness endpoints of Validator, Scan and SV apps have all been moved to /api/<app>/readyz and /api/<app>/livez, with <app> being validator, scan or sv, respectively. The corresponding Helm charts have been updated to reflect this change.

    • The URL configuration for the foundation’s Scan app in validator-values.yaml has been updated to be https://scan.sv-1.svc.TARGET_CLUSTER.network.canton.global. Similarly, in the config files in the self-hosted validator section.

    • The isDevNet flag has been removed from the cn-cometbft helm chart in order to eliminate its potential for accidental misconfiguration. Instead, the chart now relies on the value of genesis.chainId in cometbft-values.yaml to determine whether it is a TestNet or DevNet deployment.

    • The CNS UI now uses the validator API to manage user entries, instead of the JSON Ledger API. In order to authenticate to the validator-app correctly, the chart now uses the auth.audience value to specify JWT audience, instead of auth.ledgerApiAudience

  • Documentation:

    • Add documentation for configuring the SV node to publish the URL of its sequencer, so that other validators can subscribe to it.

    • Add documentation for a new required ingress rule to expose global-domain-sequencer in Configuring the Cluster Ingress.

2023-11-06

  • Deployment updates:

    • CometBFT state sync i.e. snapshot based syncing of CometBFT nodes is now enabled by default. This allows CometBFT nodes to catchup much quicker during initialization. cometbft-values.yaml is configured by default to fetch the snapshots from https://sv.sv-1.svc.TARGET_CLUSTER.network.canton.global:443/cometbft-rpc/ which is the URL for the CometBFT RPC API of the Canton-Foundation SV. In order to disable state sync, set stateSync.enable to false in cometbft-values.yaml. Further details can be found in Configuring CometBft state sync.

    • Added useSequencerConnectionsFromScan. For validator Helm charts, useSequencerConnectionsFromScan should be set to false for SV nodes. This value is in the sv-validator-values.yaml file.

  • Documentation:

  • The Scan activity and transaction history API’s now return the round for which transactions were registered. Please see the Scan OpenAPI specification. The /coin-config-for-round API can be used to lookup the holding fees for a specific round.

2023-10-30

  • Add a hidden page to the SV UI (/leader) that allows SVs to manually trigger a reelection of the SvcRules leader, as an additional safety mechanism and measure of last resort.

  • Deployment updates:

    • Add documentation for Kubernetes-Based Deployment of a Validator node

    • The requirement for url rewriting in one of the rules has been removed: https://wallet.sv.svc.<YOUR_HOSTNAME>/api/validator no longer requires rewriting (and also has been modified from /api/v0/validator to /api/validator). For example, https://wallet.sv.svc.<YOUR_HOSTNAME>/api/validator/foobar should be forwarded to http://validator-app:5003/api/validator/foobar. In the future, the other rewrite requirements will also be removed.

    • Renamed SV_WALLET_USER_ID placeholder in validator-values.yaml to OPERATOR_WALLET_USER_ID to better reflect that this value is the operator’s user in the deployment for both SVs and standalone validator nodes

  • Bugfixes:

    • Fixed an issue with fees in Scan recent activity and transaction history API where some sender and receiver fees were not reported. If a party transfers to themself, it will now be included in the Transfer receivers property in API responses (this was previously filtered out).

2023-10-23

  • Improved BFT guarantees on SVC ledger actions and mediator verdicts. Both of these are now safe as long as less than 1/3 of SVs are faulty.

  • Non-SV validators now connect to all reachable sequencers in the domain and read from them in a BFT fashion (safe as long as less than 1/3 of them are faulty). Note that for now, only the sequencers operated by the Canton Foundation are reachable and considered for that quorum. We will soon provide instructions for making the sequencers of all SV nodes reachable as well.

  • Deployment updates

    • Replaced disableAllocateLedgerApiUserParty with svValidator. For validator Helm charts, svValidator should be set to true for SV nodes. This value is in the sv-validator-values.yaml file.

    • Default volume sizes for cn-cometbft and cn-postgres increased to 240Gi each (from, respectively, 80Gi and 160Gi), as a conservative precaution.

    • The global domain now uses CometBFT instead of a Postgres-backed domain on TestNet.

    • Each global domain node now deploys both a sequencer and mediator on both DevNet and TestNet. The domain.enable flag in sv-values.yaml no longer needs to be explicitly set for DevNet (it is true by default).

    • Added scan-values.yaml, please use that when deploying the cn-scan Helm chart. The clusterUrl value is used for looking up directory entries in the scan UI.

  • Domain fees (and traffic top-ups) are now enabled on DevNet as well. This implies that explicitly setting topup.enabled to false in validator-values.yaml for DevNet is no longer required (it is always set to true).

2023-10-16

  • Frontend updates

    • The recent activity tab in Scan now looks up CNS entries for party IDs. Party IDs are shown as-is if they are not found in CNS.

  • Removed the unused “users” field from participant identities dumps.

  • A transaction history API has been added to Scan. Please see the Scan OpenAPI specification.

2023-10-09

  • Deployment updates

    • The URL of the global domain has changed to http://sequencer.sv-1.svc.<TARGET_CLUSTER>.network.canton.global:5008. This change is reflected in the values specified in participant-values.yaml, validator-values.yaml and sv-values.yaml.

  • Frontend updates

    • The “SVC Configuration” and “Canton Coin Configuration” tabs in the SV UI were renamed to, respectively, “SVC Info” and “Canton Coin Info”, to reflect the fact that they also show non-configuration information.

  • Bugfixes:

    • Fixed broken Recent Activity tab link on the scan UI.

2023-10-02

  • Deployment updates:

    • The SV name for the node operated by Digital Asset on DevNet has been updated. Note the updated SV name in approved-sv-id-values-dev.yaml: Digital-Asset.

    • The bucket configuration structure for the SV app and validator is now flattened. projectId and bucketName are now specified at the top level rather than in a config structure.

  • Frontend updates

    • The option to vote on enabled choices has been removed. This will be superseded by support for full Daml model upgrades.

2023-09-25

  • Frontend updates:

    • New history view of past SVC governance votes and actions, listing separately “Action Needed”, “In Progress”, “Planned”, “Executed” and “Rejected” votes.

      • Click on rows to review and vote on vote requests.

      • Filter vote requests by action name, requester and dates.

  • Bugfixes:

    • When configured to operate a local domain node, the SV app now waits for the local CometBFT node to fully catch up before onboarding the local sequencer and mediator nodes. This leads to clearer log messages while the SV app is initializing and improves domain performance by avoiding spam caused by local sequencer nodes that are lagging behind during initialization.

2023-09-11

  • Deployment updates:

    • Please note an SV rename from sbi to SBI-Holdings.

    • Please note an increase in database volume size from 80Gi to 160Gi.

  • Bugfixes:

    • Fixed an issue where the UIs would incorrectly include an undefined scope to requests towards an OIDC provider.

2023-09-04

  • Deployment updates:

    • Each SV node now participates in operating the global domain (devnet only)

      • A new helm chart was introduced, global-domain that will deploy a global domain node. This helm chart is currently required only on the devnet deployment. Follow the configuration instructions and the deployment instruction to also deploy the global-domain helm chart. Please note that the order of deployment of the helm charts has been updated and the helm charts must be applied in the new order.

      • Added a new value, domain.enable to the sv-values.yaml. This must be set to true only for the devnet deployment. The default value is false. Setting this to true will enable the usage of the components deployed as part of the global domain node and will make this SV node participate in operating the global domain.

  • Documentation:

    • Configuring the SV node to participate in operating the global domain (devnet only)

      • Added new entry to Configuring the Helm Charts that details the changes required for the global-domain-values.yaml

      • Included the new helm chart global-domain in the Installing the Helm Charts section and updated the helm deploy ordering

      • Added new entry to Logging into the SV UI that explains how to check if the components of the global-domain helm chart were installed correctly.

2023-08-28

  • Frontend updates:

    • The expiration of a vote request (VoteRequestTimeout) can be specified on the request level fulfilling the condition that the effective date of the coin configuration schedule change must be after the expiration.

    • Better vote requests concurrency management by preventing SVs to define coin configuration schedule effective at the same time.

  • Deployment updates:

    • Introduced a new value, isDevNet, for the CometBFT Helm charts, which should be set to true only for the devnet deployment. This value is required to generate the proper CometBFT genesis file. The value has been added to the cometbft-values.yaml file, default to false.

    • The foundingSvApiUrl has been removed from sv-values.yaml and validator-values.yaml.

  • Jfrog Artifact API Keys will be deprecated during the second half of 2023. You will need to to update the passwords of both the helm repository and the k8s secret from an API key to an Identity Token.

  • Frontend updates:

    • Retired the action to entirely replace the coin config schedule SetConfigSchedule. The governance actions to add/remove/modify individual scheduled coin config changes should be used instead.

    • Added the current CometBFT validator set to the CometBFT debug tab in the SV app

  • Documentation:

    • Clarified under which conditions it is possible to recover from a participant identities backup, and that coin balances are only preserved across TestNet resets.

2023-08-14

  • SV and validator apps now fail earlier and with a more clear error message if the version of their node software mismatches the version on the cluster they are connecting to.

  • Deployment updates:

    • The SV name for the node operated by Fiutur been updated. Note the updated SV name in approved-sv-id-values-dev.yaml: Fiutur.

  • Documentation:

    • The list of approved SV members has been moved from the sv-values.yaml file into a separate config file. Two versions thereof exist now - one for TestNet-approved identities and the other for DevNet-approved ones. The instructions for deploying the cn-sv-node Helm chart have been updated to use this separate values file.

  • The validator service now automatically uploads the directory service Daml DAR package to its participant. End-users and bootstrap scripts no longer need to upload the directory app models explicitly.

2023-08-07

  • Frontend updates:

    • Introduced new governance actions to add/remove/modify individual scheduled coin config changes. The existing UI for replacing the entire schedule will be deprecated in a coming release.

  • Deployment updates:

    • Introduced a new value, disableAllocateLedgerApiUserParty, for validator Helm charts, which should be set to true for SV nodes. This prevents clashes between the Validator and SV apps. The runbook includes this value in a newly-introduced sv-validator-values.yaml file.

    • The SV name for the node operated by IEU on behalf of LCV has been updated. Note the updated SV name in sv-values.yaml: Liberty-City-Ventures.

    • Cumberland’s name and public key have also been added in sv-values.yaml, to be included in the sv-app’s configuration as another identity that should be approved to join the SVC on TestNet.

    • Removed the SV cometbft.automationEnabled option from sv-values.yaml. The automation is enabled by default and the CometBFT network is updated to reflect the current SV network.

  • Documentation:

    • Removed obsolete section on manually “Onboarding a SV”. The subsections on generating SV and CometBFT identities have been moved into the Helm-based deployment section.

  • Bugfixes:

    • Fix an issue where after a token refresh, ledger ingestion would not get restarted.

2023-07-31

  • Frontend updates:

    • Enhanced the SetEnabledChoices Vote Request to modify the CoinRules Choices, providing increased configurability and control over the Coin.

    • Various Frontend improvements.

  • Bugfixes:

    • Fixed issue where Scan UI showed future coin config changes as if they are current.

    • Fixed issue where Scan app did not aggregate the total coin balance correctly when balance was migrated across network upgrades.

    • Fixed issue where coin migrations were not appearing in the wallet transaction log.

    • Fixed issue where automation in SV app gets into a busy loop and the app becomes irresponsive.

    • Fixed issue where existing CometBFT nodes were breaking the creation of a new network. Nodes running an old version will not be able to join new networks until they are upgraded. Note that the CometBFT Helm chart now includes the version in the genesis chain ID in order to support that.

  • SV and validator apps will now exit with an error if the version of their node software mismatches the version on the cluster they are connecting to.

  • Documentation:

    • Added a section Renaming an SV in the page “Kubernetes-Based Deployment of a Super Validator node” for step-by-step guide of renaming an SV.

2023-07-16

  • Deployment

    • Ensured that both the cn-cometbft and cn-postgres charts support the db.volumeSize and db.volumeStorageClass values for configuring persistent storage.

    • The three secrets, cn-app-scan-ledger-api-auth, cn-app-directory-ledger-api-auth, cn-app-svc-ledger-api-auth that were required before with dummy values, are no longer required.

    • The cn-postgres and cn-participant`charts now require a non-empty `postgresPassword value to be set. The value templates includes a default value that you can modify to something more secure.

    • The SV Helm runbook has been extended with a section that explains how to restore from a participant identities backup.

    • The instructions for self hosted validators have been extended with a section that explains how to restore from a participant identities backup.

    • The secrets cn-app-sv1-validator-ledger-api-auth and cn-app-sv1-ledger-api-auth are no longer required.

    • The participant now requests 4 CPUs to improve behavior under load.

  • Frontend updates:

    • Total coin balance in Scan UI now shows real data from the backend, and also reflects data from network inception and not only from the point the specific backend was created. Note that this holds only for total coin balance for now, not for all data in the Scan UI.

    • Various SV UI improvements.

2023-07-02

  • The URL to scan in validator-values.yaml has changed to https://scan.sv-1.svc.TARGET_CLUSTER.network.canton.global/api/v0/scan. The scanPort field has been removed.

  • The svSponsorAddress from validator-values.yaml has been removed. The validator associated with an SV node is onboarded automatically through its SV node.

  • The request to the IAM to acquire a token is now made using content type application/x-www-form-urlencoded instead of application/json. This matches the OAuth standard and is compatible with a wider range of IAMs. No change is required if your IAM configuration was working previously.

  • The public keys of other super validators now need to be specified in the approvedSvIdentities section in sv-values.yaml. For TestNet launch these are:

    approvedSvIdentities:
  - name: sbi
    publicKey: MFkwEwYHKoZIzj0CAQYIKoZIzj0DAQcDQgAETM+CeyHvphl9RiPDKL3vVX7F+Qo4fIhJopmgU5B7IzkwSdFic20hFB6tnAuCTU+UBjqZgh8N/h9r+CTrXMPsRg==
  - name: intellecteu-canton-da-test
    publicKey: MFkwEwYHKoZIzj0CAQYIKoZIzj0DAQcDQgAEr/iPpyuFu2U914tHyNUDuECT4/AYz9J+nLQRTC8m+95yQ6Y4Oah+Y3u3o5MK4a9D+qkoNGoG6ng0HcjA6TGKmw==
  - name: Digital-Asset
    publicKey: MFkwEwYHKoZIzj0CAQYIKoZIzj0DAQcDQgAEsRRntNkOLF2Wh7JxV0rBQPgT+SendIjFLXKUXCrLbVHqomkypHQiZP8OgFMSlByOnr81fqiUt3G36LUpg/fmgA==

  • The SV node now supports domain fee top-ups in a network where domain fees are enabled. The validator-values.yaml file contains initial recommended values for those.

  • SV and validator operators are now asked to perform backups of participant identities data, to enable continuity across network resets. This is covered in a new section of the runbook.

  • The environment variable CANTON_PARTICIPANT_USERS defining the admin users to be allocated as part of participant bootstrapping is now a simple JSON array of strings, where each element is the name of another environment variable storing the name of the user to be allocated.

  • Frontend updates:

    • Coin configuration in Scan UI now shows real data from the backend.

    • Domain Fees leaderboard in Scan UI now shows real data from the backend.

2023-06-25

  • Frontend updates:

    • The SV web UI allows operators to create vote requests to propose future coin configurations. Others can vote on these. Once the majority is reached the new configurations will be applied at due times.

    • The SV web UI allows operators to create vote requests to propose a new SV Collective configuration. Others can vote on these. Once the majority is reached the new configuration is applied.

    • The SV web UI has a new tab to display the status of your sequencer and mediator. Note that the mediator and sequencer are not yet deployed as part of the runbook so you will see an error on the status page.

  • Deployment updates:

    • The container for scan now accepts the address to the participant through the CN_APP_SCAN_PARTICIPANT_ADDRESS environment variable matching the SV app and the validator app. The helm charts set the right default so if you are using them, there is no need to configure anything.

    • The audience which defines the intended consumer of the token is now configurable in Participant, Validator app, Validator web UI, SV app, SV web UI and Directory web UI.

    • Validator apps and SV apps now configure the address of the founding SV app to allow for synchronization across nodes until all aspects of the BFT domain are robust under concurrent operations.

  • All CN apps now have /readyz and /livez endpoints which can be used for Kubernetes readiness and liveness probes.

  • Bugfixes

    • Fix a bug where the Directory UI does not connect correctly to the directory backend.

2023-06-18

  • Frontend updates:

    • Super validators can now feature and unfeature an application provider in the governance tab.

  • Deployment updates:

    • The Canton Coin Scan app is now being deployed as part of the SV node in our runbook. See instructions for deploying the cn-scan Helm chart in Installing the Software, and two new required ingress rules in Configuring the Cluster Ingress. Section Using the Canton Coin Scan UI explains the UI. Note that not all fields in the Scan UI are hooked up to fetch data in the backend yet. Ones that should work at this point are the as-of round in the top-right corner, and the Validator and App leaderboards.

    • A CometBft node is now being deployed as part of the SV node in our runbook.

    • The startup order for SV nodes has changed slightly: The SV app needs to be started before the validator app now.

    • The Canton Name Service Directory UI is now being deployed as part of the SV node in our runbook. See instructions for deploying the cn-validator Helm chart in Installing the Software, and two new required ingress rules in Configuring the Cluster Ingress. Section Using the Canton Coin Directory UI explains the UI. By searching for a name in the UI, an SV operator can register the name on Canton name service if it is not yet registered.

  • Removed the svc-client config parameter from the SV app. The SVC app is no longer used for SV onboarding and initialization.

  • Auth0 tokens used in Ledger connections are renewed 2 minutes before they expire.

  • The domain connection is now initiated by the validator and SV app instead of the participant bootstrap script which requires specifying it in the globalDomainUrl field in sv-values.yaml and validator-values.yaml.

  • Bugfixes

    • Fix a bug where the SV and validator app sometimes would stop observing ledger updates if the participant was down for a longer period of time and only recovered after a restart.

2023-06-11

  • Documentation:

    • Fixed missing service level indentation in sample istio-gateway helm chart values in the ingress installation instructions

    • Added enableHealthProbes to the participant Helm chart to allow operation on versions of Kubernetes <1.24 that do not support gRPC readiness and liveness probes.

2023-06-04

  • Deployment updates:

    • Docker images now use the same versioning scheme as helm charts: <major>.<minor>.<patch>-snapshot.<commit_date>.<number_of_commits>.0.v<commit_sha_8>

  • Frontend updates:

    • Reorganized the information tab in SV UI and included rules governing canton coin (e.g. fees) in SV UI.

    • Added support for displaying details of governance vote requests, casting a vote, and updating a casted vote.

  • Bugfixes

    • Fixed the SV onboarding URL in the Helm runbook. It must be https://sv.sv-1.svc.TARGET_CLUSTER.network.canton.global/api/v0/sv rather than https://sv-1.svc.TARGET_CLUSTER.network.canton.global/api/v0/sv.

    • Fixed an issue in last week’s release where the public/private SV keys were required in both the K8s secret and in sv-values.yaml. Now they only need to be specified through the secret.

    • Fixed how the first round for which a new SV is eligible to receive SV rewards is determined. With the fix, SVs start receiving SV rewards starting from the next round that opens after that SV has joined, i.e., an SVs will not receive SV rewards for any of the rounds that have opened before the time it has joined.

  • Deployment updates:

    • The downloaded bundle now includes sample values files for the helm charts, and the instructions have been modified to list the required user-specific configuration changes required before using them.

    • auth.jwksEndpoint value in the Helm values of the participant has been renamed to auth.jwksUrl to align with the other Helm charts, and the instructions for setting them have also been made more consistent.

2023-05-28

  • Deployment updates:

    • joinWithKeyOnboarding.keyName in sv-values.yaml has been renamed to onboardingName.

    • svSponsorPort in validator-values.yaml has been removed. The port is now included in svSponsorAddress. The default sponsor address has been changed to https://sv.sv-1.svc.TARGET_CLUSTER.network.canton.global/api/v0/sv.

    • sponsorApiPort in sv-values.yaml has been removed. The port is now included in sponsorApiUrl. The default sponsor address has been changed to https://sv.sv-1.svc.TARGET_CLUSTER.network.canton.global/api/v0/sv.

    • The SV private and public key are now stored in k8s secrets.

    • Kubernetes liveness and readiness probes are configured to probe the GRPC Health Checking Protocol of the participant node.

    • The instructions for generating your SV keys now also work on MacOS.

    • Ingress Helm charts and instructions have been rewritten to be simpler and are now based on Istio instead of Nginx. See Configuring the Cluster Ingress

  • Add new initial-coin-price-vote config option to SV app

    • for configuring an SV node to vote for a given coin price during initialization, if no coin price has been voted for by this SV node yet

    • useful for persisting coin price vote preferences across cluster (re-)deployments

  • SV UI:

    • Added support for proposing a vote on an action, currently only on removing an SV member.

  • Bugfixes

    • DA’s internal automated tests are now resilient to coin price changes allowing us to change coin price votes on DA’s SVs so votes from other SVs have an observable effect.

    • Fix SV reward collection for cases in which an SV has been offline for an extended period of time. Previously, the collection of new rewards was blocked for a potentially very long time after restarting.

2023-05-21

  • Make the Kubernetes namespace for the SV node configurable in the Helm charts (now defaulting to sv in the runbook), see deployment using Helm.

  • Features introduced to the SV UI:

    • Set your desired coin price (price per round determined using median of all coin price votes by SVs)

    • View currently open mining rounds, along with their coin prices

  • Documentation improvements

  • Bugfixes

    • Fix an issue where the validator and SV app were unable to pass the well known response from IAMs other than Auth0.

2023-05-14

  • Introduce a UI for the Super Validator operator, see SV Helm-Based Runbook.

    This UI currently allows the SV operator to see information about their SV party, and the rest of the SV collective. It allows allows the SV operator to onboard a validator by generating a validator onboarding secret (see the Self-Hosted Validator runbook for how that secret is then used by the validator operator).

  • Fix a bug where cn-node sometimes failed to start with a ClassNotFoundException.

2023-05-07

  • Add wallet UI for SV user to SV runbook. Instructions exist for deployment using Helm and local deployment (deprecated). This allows the SV operator to login to their wallet and e.g. observe SV rewards accumulating.

  • Various simplifications and extensions of SV Helm-based runbook:

    • Added instructions for setting up Auth0, and creating the corresponding k8s secrets.

    • Consolidate namespaces. Everything other than docs now resides in the sv-1 namespace.

    • Simplify the ingress setup.

2023-04-30

  • Helm chart deployment of a node connected to either TestNet or DevNet. Instructions here

  • Helm chart deployment documentation online here.

2023-04-23

  • Initial Helm chart deployment of a standalone SV node. Instructions here