Tracking: zebrad Usability

[mpguerra]

This epic issue tracks usability fixes we might want to make whilst working towards our first stable release:

Issues

• Use Display for Block in errors, logs, and spans #1158
• Design: config option limits #1591
• Investigate different designs for querying a running Zebra node #1627
• Treat TryRecvError::Closed in Inbound::poll_ready as a fatal error #1655
• Add config file comments on how to configure metric, tracing, rpc endpoints #1772
• Move finalized state database and filesystem operations to blocking threads to avoid hangs #2188
• Move the metrics recorder to a blocking thread to avoid hangs #2201
• Improve the legacy chain panic message #2394
• Make config Duration easier to write in the toml file #2847
• Hide debug-only copy-state command from users #3279
• Log a useful message when the PeerSet, peer crawler, syncer, or mempool hang #3287
• Avoid deadlocks in the crawler by doing address book updates and handshakes in separate tasks #3288
• Fan out PushTransaction requests to multiple peers #3242
• Some DNS seeders and block explorers don't show Zebra nodes #1904
• Zebra validation can be very slow on macOS, and cause very slow syncs #3042
• Implement proper zebrad health check endpoint  #1023
• Make sure RocksDB debug logs don't grow too big #1323
• Provide regular progress reports for committed blocks #1380
• Revise Zebra's logging and metrics based on a consistent design #1381
• Use git metadata and platform info in Zebra #692
• Upload Zebra crates to crates.io #1480
• Reduce disk space with RocksDB compaction and compression #1326
• Zebra stable release binaries should be linked from ZFND website #1615
• Move long README sections to the book, to make it easier for new users to get started with Zebra #1617
• Provide installation scripts for common OSes and Linux distributions for Zebra  #1619
• Run zebrad start by default #1621
• Zebra should tell users about disk space requirements #1623
• zebrad terminal user interface improvements #1624
• Leverage prometheus and grafana to allow users to customize dashboards #1625
• Investigate exposing web interface to interact with zebra #1626
• Zebra should auto-discover its own external IP address #1893
• Zebra should follow OS directory conventions better #1894
• Add logging info to zebrad's command line usage message #656
• Improve diagnostics from SerializationErrors #758
• Log an info-level message when Zebra reaches each network upgrade #970
• Make old Zebra versions eventually refuse to run #1870
• Interrupt handler does not work when a blocking task is running #1351
• Design and implement graceful shutdown for Zebra #1678
• Increase PeerSet message and peer handling test coverage #3229
• Security: Stop RocksDB threads calling unexpected code when zebrad exits #3133
• Spawn the syncer as an async task #3114
• Decide if we need shorter timeouts on PeerSet fanout requests #3136
• Stop ignoring some peers when updating the address book #3292
• Ignore some panics during shutdown using the panic handler #1677
• 2. Cache unsolicited address messages, and use them as responses #3294
• 1. Fix some address crawler timing issues #3293
• Add retries for the sapling and sprout parameter downloads #3239
• Run async tasks that access AddressBook threaded mutexes on dedicated threads #3240
• Change peer set log level when there is a large number of peers #1421

[mpguerra]

Converted to a github tracking issue

[teor2345]

Usability priorities should be based on user reports from the release candidate, we can track this using a GitHub tag.