The format is based on Keep a Changelog, and this project adheres to Semantic Versioning. See MAINTAINERS.md for instructions to keep up to date.
- Added support for "requester pays" buckets on Google Storage in url, ex:
gs://my-bucket/path?project=my-project-id
- Removed 'forked' blocks references in the logs
- Removed flags: --common-forked-blocks-store-url and --merger-prune-forked-blocks-after
- Added 'sf.firehose.v2.Fetch' endpoint
- Added
tools firehose-single-block-client
command to use this endpoint
- firehose protocol bump from v1 to v2 (see upgrade procedure below)
- Renamed 'mindreader' to 'reader' in all commands and flags
- Reading blocks from merged blocks files is now more straightforward (blocks are assumed irreversible)
- Bumped all firehose dependencies (from old versions)
- Logs are now more verbose by default output to STDERR
- Removed reader-node-merge-threshold-block-age (no more merging directly in the reader node, only the merger does it now)
- New 'ready' stat in prometheus for services
- stop mindreader and delete its state file (note the previous block, your next start block will be the block preceding the 100-block boundary ex: 12345 -> startBlock: 12299)
- stop the merger
- delete remaining "one-block-files" from your oneblock folder
- start new mindreader with flag changes and by specifying the new startBlock
- rename all 'mindreader' mentions with just 'reader' (ex:
start reader-node
and--mindreader-node-data-dir
becomes--reader-node-data-dir
- rename
--common-one-blocks-store-url
to--common-one-block-store-url
(no plural) - rename
--mindreader-node-debug-deep-mind
to--reader-node-debug-firehose-logs
- remove
--mindreader-node-discard-after-stop-num
- remove
--mindreader-node-merge-threshold-block-age
- remove
--mindreader-node-blocks-chan-capacity
- start the new merger with flag changes (...) and watch it catch with live
- rename
--common-one-blocks-store-url
to--common-one-block-store-url
(no plural) - remove
--merger-time-between-store-lookups
- remove
--merger-writers-leeway
- remove
--merger-max-one-block-operations-batch-size
- add
--merger-time-between-store-pruning
- add
--merger-time-between-store-lookups
- replace the running relayer with the new one with flag changes
- remove
--relayer-merger-addr
- remove
--relayer-buffer-size
- add
--common-one-block-store-url
- wait for it to become ready (takes some time... but it worked)
- replace the firehose with the new one with flag changes:
- remove
firehose-real-time-tolerance
- add
--common-one-block-store-url
- Renamed the
mindreader
application toreader
- Renamed
common-one-blocks-store-url
tocommon-one-block-store-url
- Renamed all the
mindreader-node-*
flags toreader-node-*