v1.1.2

• Log RPC error code when block is skipped

v1.1.1

• Fix "unable to fetch block was skipped and should not have been requested" but when an endpoint is a load balancer pointing to different nodes -- this removes an optimization that reduces the number of RPC calls by assuming that no new block will appear below the 'latest slot' of your node. If you are indeed pointing to only single solana nodes, you can use the new --optimize-single-target flag to re-enable this optimization.
• Fix startup always looking for 'first block' instead of 'cursor block', failing unnecessarily on non-archive nodes
• Add --network flag (default: mainnet) -- this flag is used only to enable a special fix around block 13334464, which you don't want to skip on testnet or devnet.

v1.1.0

• Update to firehose-core version v1.6.5.

• Added support to multiple endpoints through --endpoints flag for firesol fetch rpc.

v1.0.5

• Fixes fetching "huge" blocks --> no longer panics, increased default size and added --reader-node-line-buffer-size flag to control this
• Bumped firehose-core to version v1.5.6.

v1.0.4

• Bumped firehose-core to version v1.5.1.

• Fixed firesol fetch rpc to report fetch error(s) at least once each 30s.

v1.0.3

• Bumped firehose-core bundled binary in firehose-solana built Docker image.

v1.0.2

• Bumped firehose-core to latest version.

• Fixed some edge cases when block skips.

v1.0.1

• Fixed tools check merged-blocks default range when -r <range> is not provided to now be [0, +∞] (was previously [HEAD, +∞]).

• Fixed tools check merged-blocks to be able to run without a block range provided.

• Added API Key based authentication to tools firehose-client and tools firehose-single-block-client, specify the value through environment variable FIREHOSE_API_KEY (you can use flag --api-key-env-var to change variable's name to something else than FIREHOSE_API_KEY).

• Fixed tools check merged-blocks examples using block range (range should be specified as [<start>]?:[<end>]).

• Added --substreams-tier2-max-concurrent-requests to limit the number of concurrent requests to the tier2 Substreams service.

v1.0.0

v1.0.0

Operator notes

Important

1. All firehose processes have been removed from this binary. You will need to run this program from the firecore binary
   • Previous firesol start ... command becomes firecore start ...
2. New Poller: firesol no longer gets blocks from a Bigtable instance: it fetches the blocks using RPC calls
   • Run firecore start reader with --reader-node-path=/path/to/firesol and --reader-node-arguments=fetch rpc <https://your.solana.rpc/path> <start-block>
3. New Block Format requires either fetching all the merged blocks again or converting them
   • Convert old blocks by running: ACCEPT_SOLANA_LEGACY_BLOCK_FORMAT=true firesol upgrade-merged-blocks <source-store> <dest-store> <start-num:stop-num>
4. Upgrading your deployment will require a "stop the world" upgrade, where you start the new binaries, pointing to the new blocks, without any contact with the previous blocks or components.

Removed

• All the firesol start ... commands have been removed. Use firecore binary to run the reader, merger, relayer, firehose and substreams services
• All the existing firesol tools commands

Added

• Added fetch rpc <endpoint> <start_block> command fetches and prints the blocks in protobuf format, to be used by the firecore start reader command.
• Added upgrade-merged-blocks command to perform the upgrade on previous solana merged-blocks.
• Bumped firecore version to v1.2.0

Fixed

• Fixed Substreams scheduler sometimes taking a long time to spawn more than a single worker.

v0.2.7

Operators

Important

We have had reports of older versions of this software creating corrupted merged-blocks-files (with duplicate or extra out-of-bound blocks)
This release adds additional validation of merged-blocks to prevent serving duplicate blocks from the firehose or substreams service.
This may cause service outage if you have produced those blocks or downloaded them from another party who was affected by this bug.

1. Find the affected files by running the following command (can be run multiple times in parallel, over smaller ranges)

tools check merged-blocks-batch <merged-blocks-store> <start> <stop>

2. If you see any affected range, produce fixed merged-blocks files with the following command, on each range:

tools fix-bloated-merged-blocks <merged-blocks-store> <output-store> <start>:<stop>

3. Copy the merged-blocks files created in output-store over to the your merged-blocks-store, replacing the corrupted files.

Added

• Firehose logs now include auth information (userID, keyID, realIP) along with blocks + egress bytes sent.
• Added tools check merged-blocks-batch to simplify checking blocks continuity in batched mode, optionally writing results to a store
• Added the command tools fix-bloated-merged-blocks to try to fix merged-blocks that contain duplicates and blocks outside of their range.
• Command tools print one-block and merged-blocks now supports a new --output-format jsonl format. Bytes data can now printed as hex or base58 string instead of base64 string.
• Added retry loop for merger when walking one block files. Some use-cases where the bundle reader was sending files too fast and the merger was not waiting to accumulate enough files to start bundling merged files

Fixed

• Bumped bstream: the filesource will now refuse to read blocks from a merged-files if they are not ordered or if there are any duplicate.
• The command tools download-from-firehose will now fail if it is being served blocks "out of order", to prevent any corrupted merged-blocks from being created.
• The command tools print merged-blocks did not print the whole merged-blocks file, the arguments were confusing: now it will parse <start_block> as a uint64.
• The command tools unmerge-blocks did not cover the whole given range, now fixed

Removed

• Breaking The reader-node-log-to-zap flag has been removed. This was a source of confusion for operators reporting Firehose on bugs because the node's logs where merged within normal Firehose on logs and it was not super obvious.

  Now, logs from the node will be printed to stdout unformatted exactly like presented by the chain. Filtering of such logs must now be delegated to the node's implementation and how it deals depends on the node's binary. Refer to it to determine how you can tweak the logging verbosity emitted by the node.

• Flag substreams-rpc-endpoints removed, this was present by mistake and unused actually.

• Flag substreams-rpc-cache-store-url removed, this was present by mistake and unused actually.

• Flag substreams-rpc-cache-chunk-size removed, this was present by mistake and unused actually.

(Note: release 0.2.6 was never actually released)