Version 2.2.4.1

This is a required release for users of Nethermind on Gnosis, and optional for all others

• Let Nethermind determine sync mode based on chain
• Better Nethermind prune. Thanks to Joe at RocketPool for the suggestion!
• Some more cleanup around the removed Prometheus alert manager
• Remove Akula 😭
• Fix an error introduced by shell linting that caused ./ethd terminate to fail - thanks to @RomanS-re!

Version 2.2.4

This is a required release for users of Nethermind, and optional for all others

• Support Nethermind's new engine port parameters
• ./ethd attach-geth command, thanks to @ldub!
• Removed Prometheus alert manager - it had been broken since before merge, may as well remove it 😅. Use Grafana's built-in alert manager instead, please
• Sample config file for Grafana Cloud in prometheus/custom-prom.yml.sample
• Require 250GiB free disk space before starting Nethermind prune

Version 2.2.3.1

This is a required release for users of Akula, and optional for all others

• Fixed Akula source build
• Lint pass on all shell scripts
• Added deprecation warning to ./ethd keyimport
• Prometheus now uses default scrape time, so that custom-prom.yml will work with grafana-cloud.yml

Version 2.2.3

This is an optional release containing new features

• Work around Lodestar's non-standard return codes on recipient/gas keymanager API
• Keymanager API key import now waits up to 60s for slashing protection DB to be imported
• Support Nethermind health checks
• Use beaconcha.in's stat collector for Prysm and Nimbus

Version 2.2.2

This is a required release for users of Erigon, and optional for all others

• Support Erigon 2.30.0's new --externalcl parameter
• Fix test for custom-prom.yml used in grafana-cloud.yml
• Support MEV boost -minbid
• Check for 125 GiB free disk space before Nethermind prune
• Dynamic wait time when re-starting Nethermind for prune

Version 2.2.1

This is a required release for users of Erigon, and optional for all others

• Support Erigon 2.29.0's new log level parameter
• More debug info for sporadically failing Nethermind prune

Version 2.2.0

This is a recommended release with new features and bug fixes

• Nethermind background pruning via ./ethd prune-nethermind
• Bonsai snapshots supported with Besu
• New grafana-cloud.yml that uses prometheus/custom-prom.yml. Undocumented for now
• Do not default to --subscribe-all-subnets for cl-only.yml files. Please add this via CL_EXTRAS if you need it
• Geth always runs an apk update first when building the local image
• Fix provisioning of the prysm_less_10 dashboard. Thanks @FloatingUpstream!
• Support Teku's validator exit command for API managed keys
• Better UX for initial Blox SSV setup
• Add 30d retention period to Loki
• Moved CL_EXTRAS, EL_EXTRAS and VC-EXTRAS into entrypoint script
• slasher.yml files removed. If you used these, please use CL_EXTRAS instead
• Enable pprof on Erigon and optimize for ZFS storage
• cadvisor does housekeeping every 30s
• Manifold removed from default relay list
• Nimbus VC offered during ./ethd config

Version 2.1.3

This is an optional release with new features

• Erigon uses stable docker tag
• Erigon uses --batchSize 64m in an attempt to squeeze it into 16 GiB 🐭
• Besu defaults to latest-openjdk-latest and uses soft heap 3g
• Add Nimbus vc-only yml
• Erigon and Nimbus source builds default to latest release
• Remove all OVERRIDE_TTD mentions

Version 2.1.2

This is an optional release with bug fixes

• Prometheus now survives restarts 😅
• Quiet Nethermind push failures
• Fix ext-network.yml version

Version 2.1.1

This is an optional release with new features and bug fixes

• Added BESU_HEAP and TEKU_HEAP variables to override the default Java heap settings for each
• Fixed a bug in the new Prometheus yml handling for Nimbus and Teku
• Default Besu to 5g heap, up from 4g
• Source builds use Go 1.19
• Add new dependency for Lighthouse source build
• ./ethd install installs docker-ce instead of docker.io
• ./ethd config offers previous values to the user
• ELs no longer initial-start as root; all EL and CL will chown their directories and then restart with their non-root user if started as root
• ./ethd update warns the user if there are uncommitted local changes