v0.36.0

v0.36.0

Note on Firehose Extended Block Details

By default, all Firehose providers are required to support extended block details, as this is the safest option for a graph-node operator. Firehose providers that do not support extended block details for enabled chains are considered invalid and will not be used.

To disable checks for one or more chains, simply specify their names
in GRAPH_NODE_FIREHOSE_DISABLE_EXTENDED_BLOCKS_FOR_CHAINS as a comma separated list of chain names. Graph Node defaults to an empty list, which means that this feature is enabled for all chains.

What's new

• Add support for substreams using 'index modules', 'block filters', 'store:sum_set'. (#5463)
• Implement new IPFS client (#5600)
• Add timestamp support to substreams. (#5641)
• Add graph-indexed header to query responses. (#5710)
• Use the new Firehose info endpoint. (#5672)
• Store synced_at_block_number when a deployment syncs. (#5610)
• Create nightly docker builds from master branch. (#5400)
• Make sure transact_block_operations does not go backwards. (#5419)
• Improve error message when store write fails. (#5420)
• Allow generating map of section nesting in debug builds. (#5279)
• Ensure substream module name is valid. (#5424)
• Improve error message when resolving references. (#5385)
• Check if subgraph head exists before trying to unfail. (#5409)
• Check for EIP 1898 support when checking block receipts support. (#5406)
• Use latest block hash for check_block_receipts. (#5427)
• Handle null blocks from Lotus. (#5294)
• Increase firehose grpc max decode size. (#5483)
• Improve Environment variable docs, rename GRAPH_ETHEREUM_BLOCK_RECEIPTS_TIMEOUT to GRAPH_ETHEREUM_BLOCK_RECEIPTS_CHECK_TIMEOUT. (#5468)
• Remove provider checks at startup. (#5337)
• Track more features in subgraph features table. (#5479)
• Implement is_duplicate_of for substreams. (#5482)
• Add docs for GRAPH_POSTPONE_ATTRIBUTE_INDEX_CREATION. (#5515)
• Improve error message for missing template during grafting. (#5464)
• Enable "hard-coded" values in declarative eth_calls. (#5498)
• Respect causality region in derived fields. (#5488)
• Improve net_identifiers call with timeout. (#5549)
• Add arbitrum-sepolia chain ID to GRAPH_ETH_CALL_NO_GAS default value. (#5504)
• Disable genesis validation by default. (#5565)
• Timeout when trying to get net_identifiers at startup. (#5568)
• Only start substreams if no other block investor is available. (#5569)
• Allow running a single test case for integration tests. (#5577)
• Store timestamp when marking subgraph as synced. (#5566)
• Document missing env vars. (#5580)
• Return more features in status API. (#5582)
• Respect substreams datasource startBlock. (#5617)
• Update flagged dependencies. (#5659)
• Add more debug logs when subgraph is marked unhealthy. (#5662)
• Add config option for cache stores. (#5716)

Bug fixes

• Add safety check when rewinding. (#5423)
• Fix rewind for deployments with multiple names. (#5502)
• Improve graphman copy performance (#5425)
• Fix retrieving chain info with graphman for some edge cases. (#5516)
• Improve graphman restart to handle multiple subgraph names for a deployment. (#5674)
• Improve adapter startup. (#5503)
• Detect Nethermind eth_call reverts. (#5533)
• Fix genesis block fetching for substreams. (#5548)
• Fix subgraph_resume being mislabelled as pause. (#5588)
• Make SubgraphIndexingStatus.paused nullable. (#5551)
• Fix a count aggregation bug. (#5639)
• Fix prost generated file. (#5450)
• Fix deployment_head metrics not progressing for substreams. (#5522)
• Enable graft validation checks in debug builds. (#5584)
• Use correct store when loading indexes for graft base. (#5616)
• Sanitise columns in SQL. (#5578)
• Truncate subgraph_features table before migrating. (#5505)
• Consistently apply max decode size. (#5520)
• Various docker packaging improvements (#5709) (#5711) (#5712) (#5620) (#5621)
• Retry IPFS requests on Cloudflare 521 Web Server Down. (#5687)
• Optimize IPFS retries. (#5698)
• Exclude full-text search columns from entity queries. (#5693)
• Do not allow multiple active runners for a subgraph. (#5715)
• Stop subgraphs passing max endBlock. (#5583)
• Do not repeat a rollup after restart in some corner cases. (#5675)

Graphman

• Add command to update genesis block for a chain and to check genesis information against all providers. (#5517)
• Create GraphQL API to execute commands (#5554)
• Add graphman create/remove commands to GraphQL API. (#5685)

Contributors

Thanks to all contributors for this release: @dwerner, @encalypto, @incrypto32, @isum, @leoyvens, @lutter, @mangas, @sduchesneau, @shiyasmohd, @shuaibbapputty, @YaroShkvorets, @ziyadonji, @zorancv

Full Changelog: v0.35.1...v0.36.0

v0.35.1

New subgraph features

• Declarative Eth calls - support executing Eth calls ahead of time and in parallel. Using this feature, the time for all calls goes from the sum of the time the calls take to the max. #5264
• Support for differentiating EOA from smart contract accounts - ethereum.getCode can be used in mappings to check whether smart contract code is present at a given address. #5272
• Support additional event filters - allow filtering on Indexed arguments. #5254
• Aggregations now use Timestamp type for the timestamp attribute. #5321

Optimizations

• Fetch block receipts in bulk - use eth_getBlockReceipts on supported providers to fetch receipts in bulk. #5341
• Improve indices created to support aggregations. #5286
• Use 'select by specific attributes' by default. #5311
• Select only the needed specific attributes for query_window_one_entity. #5357
• Check interface uniqueness in bulk. #5293
• Stop using the IPFS stat API in favor of the more efficient cat API. #5284
• Derive macros for CheapClone and CacheWeight. #5326
• Periodically remove entries from the LayoutCache. #5324
• Centralize and clarify our use of futures. #5345
• Optimize subgraph ptr update. #5317

Graphman

• Refactor manual index creation & add ddl tests. #5292
• Refactor graphman rewind to use pause and resume logic. #5213

Firehose & Substreams

• Add deployment id to Firehose connection headers. #5319
• Improve the Ethereum Firehose codec. #5314
• Fix static filters restart behaviour when using threshold. #5328
• Try to be more consistent in Firehose parsing to address. #5360
• Include 'provider' in various firehose log messages. #5384
• Do not send all Firehose headers when its init triggers only. #5373

Robustness

• Add a test for number_gte block constraints. #5383
• Add zkSync fingerprint to the TOO_MANY_LOGS_FINGERPRINTS. #5307
• Add exceptional case for transaction less event in zksync-era. #5338
• Improve tests by moving subgraph param to TestContext. #5342

Notable dependency upgrades

• Upgrade clap. #5174
• Update syn to 2.0. #5339
• Update tokio-tungstenite. #5344

Bug fixes

• Avoid 'too many bind params' error in FindDerivedEntityQuery. #5318
• Fix check for deterministic error. #5332
• Fix incorrect comparison of transaction hash. #5374
• Fix clap. #5389

Contributors

• @itsjerryokolo made their first contribution in #5213! Thanks for the contribution @itsjerryokolo!🙌
• Thanks to all contributors to this release: @incrypto32, @lutter, @leoyvens, @mangas, @itsjerryokolo, @YaroShkvorets, @fordN, and of course the trusty @dependabot.

For readability not all commits and changes were noted here. To see a complete list head to the full changelog at v0.35.0...v0.35.1

v0.35.0

What's new

• Aggregations - Declarative aggregations defined in the subgraph schema allow the developer to aggregate values on specific intervals using flexible aggregation functions. (#5082) (#5184) (#5209) (#5242) (#5208)
• Add pause and resume to admin JSON-RPC API - Adds support for explicit pausing and resuming of subgraph deployments with a field tracking the paused state in indexerStatuses. (#5190)
• Support eth_getBalance calls in subgraph mappings - Enables fetching the Eth balance of an address from the mappings using ethereum.getBalance(address). (#5202)
• Add parentHash to _meta query - Particularly useful when polling for data each block to verify the sequence of blocks. (#5232)
• Parallel execution of all top-level queries in a single query body (#5273)
• The ElasticSearch index to which graph-node logs can now be configured with the GRAPH_ELASTIC_SEARCH_INDEX environment variable which defaults to subgraph. (#5210)
• Some small prefetch simplifications. (#5132)
• Migration changing the type of health column to text. (#5077)
• Disable eth_call_execution_time metric by default. (#5164)
• Call revert_state_to whenever blockstream is restarted. (#5187)
• Pruning performance improvement: only analyze when rebuilding. (#5186)
• Disallow grafts within the reorg threshold. (#5135)
• Optimize subgraph synced check-less. (#5198)
• Improve error log. (#5217)
• Update provider docs. (#5216)
• Downgrade 'Entity cache statistics' log to trace. (#5241)
• Do not clone MappingEventHandlers in match_and_decode. (#5244)
• Make batching conditional on caught-up status. (#5252)
• Remove hack in chain_head_listener. (#5240)
• Increase sleep time in write queue processing. (#5266)
• Memoize Batch.indirect_weight. (#5276)
• Optionally track detailed indexing gas metrics in csv. (#5215)
• Do not use prefix comparisons for primary keys. (#5289)

Graphman

• Add ability to list removed unused deployment by id. (#5152)
• Add command to change block cache shard. (#5169)

Firehose and Substreams

• Add key-based authentication for Firehose/Substreams providers. (#5259)
• Increase blockstream buffer size for substreams. (#5182)
• Improve substreams error handling. (#5160)
• Reset substreams/firehose block ingestor backoff. (#5047)

Bug Fixes

• Fix graphiql issue when querying subgraph names with multiple path segments. (#5136)
• Fix change_health_column migration for sharded setup. (#5183)
• Fix conversion of BlockTime for NEAR. (#5206)
• Call revert_state_to to last good block instead of current block. (#5195)
• Fix Action::block_finished. (#5218)
• Fix runtime timeouts. (#5236)
• Remove panic from rewind and truncate. (#5233)
• Fix version stats for huge number of versions. (#5261)
• Fix _meta query failure due to incorrect selection set use. (#5265)

Major dependency upgrades

• Update to diesel v2. (#5002)
• Bump Rust version. (#4985)

Contributors

Thank you to all the contributors! @incrypto32, @mangas, @lutter, @leoyvens, @zorancv, @YaroShkvorets, @seem-less

Full Changelog: v0.34.1...v0.35.0

v0.34.1

$ docker pull graphprotocol/graph-node:v0.34.1

This a patch release that addresses a few issues

• Fixes an issue that caused an increase in data size of /metrics endpoint of graph-node. (#5161). Indexers are advised to skip v0.34.0 and use v0.34.1
• Fixes an issue that caused subgraphs with file data sources to skip non-deterministic errors that occurred in a file data source mapping handler.

v0.34.0

$ docker pull graphprotocol/graph-node:v0.34.0

What's New

• Substreams as Source of Triggers for Subgraphs - This update significantly enhances subgraph functionality by enabling substreams to act as a source of triggers for running subgraph mappings. Developers can now directly run subgraph mappings on the data output from substreams, facilitating a more integrated and efficient workflow.(#4887) (#4916)
• indexerHints in Manifest for Automated Pruning - This update introduces the ability for subgraph authors to specify indexerHints with a field prune in their manifest, indicating the desired extent of historical block data retention. This feature enables graph-node to automatically prune subgraphs when the stored history exceeds the specified limit, significantly improving query performance. This automated process eliminates the need for manual action by indexers for each subgraph. Indexers can also override user-set historyBlocks with the environment variable GRAPH_HISTORY_BLOCKS_OVERRIDE (#5032 (#5117)
• Initial Starknet Support - Introducing initial Starknet support for graph-node, expanding indexing capabilities to the Starknet ecosystem. The current integration is in its early stages, with notable areas for development including the implementation of trigger filters and data source template support. Future updates will also bring substream support. (#4895)
• endBlock Feature in Data Sources - This update adds the endBlock field for dataSources in subgraph manifest. By setting an endBlock, subgraph authors can define the exact block at which a data source will cease processing, ensuring no further triggers are processed beyond this point. (#4787
• Autogenerated Int8 IDs in graph-node - Introduced support for using Int8 as the ID type for entities, with the added capability to auto-generate these IDs, enhancing flexibility and functionality in entity management. (#5029)
• GraphiQL V2 Update - Updated GraphiQL query interface of graph-node to version 2. (#4677)
• Sharding Guide for Graph-Node - A new guide has been added to graph-node documentation, explaining how to scale graph-node installations using sharding with multiple Postgres instances. Sharding Guide
• Per-chain polling interval configuration for RPC Block Ingestors (#5066)
• Metrics Enhancements(#5055) (#4937)
• graph-node now avoids creating GIN indexes on array attributes to enhance database write performance, addressing the issue of expensive updates and underutilization in queries. (#4933)
• The subgraphFeatures endpoint in graph-node has been updated to load features from subgraphs prior to their deployment. (#4864)
• Improved log filtering performance in blockstream. (#5015)
• Enhanced GraphQL error reporting by including __schema and __type fields in the results during indexing errors (#4968)

Bug fixes

• Addressed a bug in the deduplication logic for Cosmos events, ensuring all distinct events are properly indexed and handled, especially when similar but not identical events occur within the same block. (#5112)
• Fixed compatibility issues with ElasticSearch 8.X, ensuring proper log functionality. (#5013)
• Resolved an issue when rewinding data sources across multiple blocks. In rare cases, when a subgraph had been rewound by multiple blocks, data sources 'from the future' could have been left behind. This release adds a database migration that fixes that. With very unlucky timing this migration might miss some subgraphs, which will later lead to an error assertion failed: self.hosts.last().and_then(|h| h.creation_block_number()) <= data_source.creation_block(). Should that happen, the migration script should be rerun against the affected shard. (#5083)
• Increased the base backoff time for RPC, enhancing stability and reliability under load. (#4984)
• Resolved an issue related to spawning offchain data sources from existing offchain data source mappings. (#5051)(#5092)
• Resolved an issue where eth-call results for reverted calls were being cached in call cache. (#4879)
• Fixed a bug in graphman's index creation to ensure entire String and Bytes columns are indexed rather than just their prefixes, resulting in optimized query performance and accuracy. (#4995)
• Adjusted SubstreamsBlockIngestor to initiate at the chain's head block instead of starting at block zero when no cursor exists. (#4951)
• Fixed a bug that caused incorrect progress reporting when copying subgraphs, ensuring accurate status updates. (#5075)

Graphman

• Graphman Deploy Command - A new graphman deploy command has been introduced, simplifying the process of deploying subgraphs to graph-node. (#4930)

Complete changelog: v0.33.0...v0.34.0

v0.34.0-rc.0

$ docker pull graphprotocol/graph-node:v0.34.0-rc.0

What's New

• Substreams as Source of Triggers for Subgraphs - This update significantly enhances subgraph functionality by enabling substreams to act as a source of triggers for running subgraph mappings. Developers can now directly run subgraph mappings on the data output from substreams, facilitating a more integrated and efficient workflow.(#4887) (#4916)
• indexerHints in Manifest for Automated Pruning - This update introduces the ability for subgraph authors to specify indexerHints with a field prune in their manifest, indicating the desired extent of historical block data retention. This feature enables graph-node to automatically prune subgraphs when the stored history exceeds the specified limit, significantly improving query performance. This automated process eliminates the need for manual action by indexers for each subgraph. Indexers can also override user-set historyBlocks with the environment variable GRAPH_HISTORY_BLOCKS_OVERRIDE (#5032 (#5117)
• Initial Starknet Support - Introducing initial Starknet support for graph-node, expanding indexing capabilities to the Starknet ecosystem. The current integration is in its early stages, with notable areas for development including the implementation of trigger filters and data source template support. Future updates will also bring substream support. (#4895)
• endBlock Feature in Data Sources - This update adds the endBlock field for dataSources in subgraph manifest. By setting an endBlock, subgraph authors can define the exact block at which a data source will cease processing, ensuring no further triggers are processed beyond this point. (#4787
• Autogenerated Int8 IDs in graph-node - Introduced support for using Int8 as the ID type for entities, with the added capability to auto-generate these IDs, enhancing flexibility and functionality in entity management. (#5029)
• GraphiQL V2 Update - Updated GraphiQL query interface of graph-node to version 2. (#4677)
• Sharding Guide for Graph-Node - A new guide has been added to graph-node documentation, explaining how to scale graph-node installations using sharding with multiple Postgres instances. Sharding Guide
• Per-chain polling interval configuration for RPC Block Ingestors (#5066)
• Metrics Enhancements(#5055) (#4937)
• graph-node now avoids creating GIN indexes on array attributes to enhance database write performance, addressing the issue of expensive updates and underutilization in queries. (#4933)
• The subgraphFeatures endpoint in graph-node has been updated to load features from subgraphs prior to their deployment. (#4864)
• Improved log filtering performance in blockstream. (#5015)
• Enhanced GraphQL error reporting by including __schema and __type fields in the results during indexing errors (#4968)

Bug fixes

• Addressed a bug in the deduplication logic for Cosmos events, ensuring all distinct events are properly indexed and handled, especially when similar but not identical events occur within the same block. (#5112)
• Fixed compatibility issues with ElasticSearch 8.X, ensuring proper log functionality. (#5013)
• Resolved an issue when rewinding data sources across multiple blocks. In rare cases, when a subgraph had been rewound by multiple blocks, data sources 'from the future' could have been left behind. This release adds a database migration that fixes that. With very unlucky timing this migration might miss some subgraphs, which will later lead to an error assertion failed: self.hosts.last().and_then(|h| h.creation_block_number()) <= data_source.creation_block(). Should that happen, the migration script should be rerun against the affected shard. (#5083)
• Increased the base backoff time for RPC, enhancing stability and reliability under load. (#4984)
• Resolved an issue related to spawning offchain data sources from existing offchain data source mappings. (#5051)(#5092)
• Resolved an issue where eth-call results for reverted calls were being cached in call cache. (#4879)
• Fixed a bug in graphman's index creation to ensure entire String and Bytes columns are indexed rather than just their prefixes, resulting in optimized query performance and accuracy. (#4995)
• Adjusted SubstreamsBlockIngestor to initiate at the chain's head block instead of starting at block zero when no cursor exists. (#4951)
• Fixed a bug that caused incorrect progress reporting when copying subgraphs, ensuring accurate status updates. (#5075)

Graphman

• Graphman Deploy Command - A new graphman deploy command has been introduced, simplifying the process of deploying subgraphs to graph-node. (#4930)

v0.33.0

What's New

• Arweave file data sources - Arweave file data sources allow subgraph developers to access offchain data from Arweave from within the subgraph mappings.(#4789)
• Major performance boost for substreams-based subgraphs - Significant performance improvements have been achieved for substreams-based subgraphs by moving substreams processing to the block stream.(#4851)
• Polling block handler - A new block handler filter polling for ethereum data sources which enables subgraph developers to run a block handler at defined block intervals. This is useful for use cases such as taking periodic snapshots of the contract state.(#4725)
• Initialization handler - A new block handler filter once for ethereum data sources which enables subgraph developers to create a handler which will be called only once before all other handlers run. This configuration allows the subgraph to use the handler as an initialization handler, performing specific tasks at the start of indexing. (#4725)
• DataSourceContext in manifest - DataSourceContext in Manifest - DataSourceContext can now be defined in the subgraph manifest. It's a free-form map accessible from the mapping. This feature is useful for templating chain-specific data in subgraphs that use the same codebase across multiple chains.(#4848)
• graph-node version in index node API - The Index Node API now features a new query, Version, which can be used to query the current graph-node version and commit. (#4852)
• Added a 'paused' field to Index Node API, a boolean indicating the subgraph’s pause status. (#4779)
• Proof of Indexing logs now include block number (#4798)
• subgraph_features table now tracks details about handlers used in a subgraph (#4820)
• Configurable SSL for Postgres in Dockerfile - ssl-mode for Postgres can now be configured via the connection string when deploying through Docker, offering enhanced flexibility in database security settings.(#4840)
• Introspection Schema Update - The introspection schema has been updated to align with the October 2021 GraphQL specification update.(#4676)
• trace_id Added to Substreams Logger (#4868)
• New apiVersion for Mapping Validation - The latest apiVersion 0.0.8 validates that fields set in entities from the mappings are actually defined in the schema. This fixes a source of non-deterministic PoI. Subgraphs using this new API version will fail if they try to set undefined schema fields in the mappings. Its strongly recommended updating to 0.0.8 to avoid these issues. (#4894)
• Substreams Block Ingestor Support - Added the ability to run a pure substreams chain by introducing a block ingestor for substreams-only chains. This feature allows users to run a chain with just a single substreams endpoint, enhancing support beyond RPC and firehose. Prior to this, a pure substreams chain couldn’t be synced.(#4839)

Bug fixes

• Fix for rewinding dynamic data source - Resolved an issue where a rewind would fail to properly remove dynamic data sources when using graphman rewind. This has been fixed to ensure correct behavior.(#4810)
• Improved Deployment Reliability with Retry Mechanism - A retry feature has been added to the block_pointer_from_number function to enhance the robustness of subgraph deployments. This resolves occasional failures encountered during deployment processes.(#4812)
• Fixed Cross-Shard Grafting Issue - Addressed a bug that prevented cross-shard grafting from starting, causing the copy operation to stall at 0% progress. This issue occurred when a new shard was added after the primary shard had already been configured. The fix ensures that foreign tables and schemas are correctly set up in new shards. For existing installations experiencing this issue, it can be resolved by running graphman database remap.(#4845)
• Fixed a Full-text search regression - Reverted a previous commit (ad1c6ea) that inadvertently limited the number of populated search indexes per entity.(#4808)
• Attestable Error for Nested Child Filters - Nested child filter queries now return an attestable ChildFilterNestingNotSupportedError, improving error reporting for users.(#4828)

Graphman

• Index on prefixed fields - The graphman index create command now correctly indexes prefixed fields of type String and Bytes for more query-efficient combined indexes. Note: For fields that are references to entities, the behavior may differ. The command may create an index using left(..) when it should index the column directly.
• Partial Indexing for Recent Blocks - The graphman index create command now includes a --after $recent_block flag for creating partial indexes focused on recent blocks. This enhances query performance similar to the effects of pruning. Queries using these partial indexes must include a specific clause for optimal performance.(#4830)

Full Changelog: incrypto32/v0.32.0...v0.33.0

v0.33.0-rc.0

$ docker pull graphprotocol/graph-node:v0.33.0-rc.0

What's New

• Arweave file data sources - Arweave file data sources allow subgraph developers to access offchain data from Arweave from within the subgraph mappings.(#4789)
• Major performance boost for substreams-based subgraphs - Significant performance improvements have been achieved for substreams-based subgraphs by moving substreams processing to the block stream.(#4851)
• Polling block handler - A new block handler filter polling for ethereum data sources which enables subgraph developers to run a block handler at defined block intervals. This is useful for use cases such as taking periodic snapshots of the contract state.(#4725)
• Initialization handler - A new block handler filter once for ethereum data sources which enables subgraph developers to create a handle which will be called only once before all other handlers run. This configuration allows the subgraph to use the handler as an initialization handler, performing specific tasks at the start of indexing. (#4725)
• DataSourceContext in manifest - DataSourceContext in Manifest - DataSourceContext can now be defined in the subgraph manifest. It's a free-form map accessible from the mapping. This feature is useful for templating chain-specific data in subgraphs that use the same codebase across multiple chains.(#4848)
• graph-node version in index node API - The Index Node API now features a new query, Version, which can be used to query the current graph-node version and commit. (#4852)
• Added a 'paused' field to Index Node API, a boolean indicating the subgraph’s pause status. (#4779)
• Proof of Indexing logs now include block number (#4798)
• subgraph_features table now tracks details about handlers used in a subgraph (#4820)
• Configurable SSL for Postgres in Dockerfile - ssl-mode for Postgres can now be configured via the connection string when deploying through Docker, offering enhanced flexibility in database security settings.(#4840)
• Introspection Schema Update - The introspection schema has been updated to align with the October 2021 GraphQL specification update.(#4676)
• trace_id Added to Substreams Logger (#4868)
• New apiVersion for Mapping Validation - The latest apiVersion 0.0.8 now validates that fields set in entities from the mappings are actually defined in the schema. Subgraphs using this new API version will fail if they try to set undefined schema fields in the mappings. (#4894)
• Substreams Block Ingestor Support - Added the ability to run a pure substreams chain by introducing a block ingestor for substreams-only chains. This feature allows users to run a chain with just a single substreams endpoint, enhancing support beyond RPC and firehose. Prior to this, a pure substreams chain couldn’t be synced.(#4839)

Bug fixes

• Fix for rewinding dynamic data source - Resolved an issue where a rewind would fail to properly remove dynamic data sources when using graphman rewind. This has been fixed to ensure correct behavior.(#4810)
• Improved Deployment Reliability with Retry Mechanism - A retry feature has been added to the block_pointer_from_number function to enhance the robustness of subgraph deployments. This resolves occasional failures encountered during deployment processes.(#4812)
• Fixed Cross-Shard Grafting Issue - Addressed a bug that prevented cross-shard grafting from starting, causing the copy operation to stall at 0% progress. This issue occurred when a new shard was added after the primary shard had already been configured. The fix ensures that foreign tables and schemas are correctly set up in new shards.(#4845)
• Fixed a Full-text search regression - Reverted a previous commit (ad1c6ea) that inadvertently limited the number of populated search indexes per entity.(#4808)
• Attestable Error for Nested Child Filters - Nested child filter queries now return an attestable ChildFilterNestingNotSupportedError, improving error reporting for users.(#4828)

Graphman

• Index on prefixed fields - The graphman index create command now correctly indexes prefixed fields of type String and Bytes, making combined indexes more query-efficient.
• Partial Indexing for Recent Blocks - The graphman index create command now includes a --after $recent_block flag for creating partial indexes focused on recent blocks. This enhances query performance similar to the effects of pruning. Queries using these partial indexes must include a specific clause for optimal performance.(#4830)

Full Changelog: v0.32.0...b1502a4

v0.32.0

$ docker pull graphprotocol/graph-node:v0.32.0

What's New

• Derived fields getter: Derived fields can now be accessed from within the mapping code during indexing. (#4434)
• Sorting interfaces by child entity: Interfaces can now be sorted by non-derived child entities. (#4058)
• File data sources can now be spawned from handlers of other file data sources: This enables the use of file data sources for scenarios where a file data source needs to be spawned from another one. One practical application of this feature is in handling NFT metadata. In such cases, the metadata itself is stored as a file on IPFS and contains embedded IPFS CID for the actual file for the NFT. (#4713)
• Allow redeployment of grafted subgraphs even when graft_base is not available: This will allow renaming of already synced grafted subgraphs even when the graft base is not available, which previously failed due to graft-base validation errors. (#4695)
• history_blocks is now available in the index-node API. (#4662)
• Added a new subgraph features table in primary to easily track information like apiVersion, specVersion, features, and data source kinds used by subgraphs. (#4679)
• subgraphFeatures endpoint now includes data from subgraph_features table.
• ens_name_by_hash is now undeprecated: This reintroduces support for fetching ENS names by their hash, dependent on the availability of the underlying Rainbow Table (#4751).
• Deterministically failed subgraphs now return valid POIs for subsequent blocks after the block at which it failed. (#4774)
• eth-call logs now include block hash and block number: This enables easier debugging of eth-call issues. (#4718)
• Enabled support for substreams on already supported networks. (#4767)
• Add new GraphQL scalar type Int8. This new scalar type allows subgraph developers to represent 8-bit signed integers. (#4511)
• Add support for overriding module params for substreams-based subgraphs when params are provided in the subgraph manifest. (#4759)

Breaking changes

• Duplicate provider labels are not allowed in graph-node config anymore

Bug fixes

• Fixed PublicProofsOfIndexing returning the error Null value resolved for non-null field proofOfIndexing when fetching POIs for blocks that are not in the cache (#4768)
• Fixed an issue where Block stream would fail when switching back to an RPC-based block ingestor from a Firehose ingestor. (#4790)
• Fixed an issue where derived loaders were not working with entities with Bytes as IDs (#4773)
• Firehose connection test now retries for 30 secs before setting the provider status to Broken (#4754)
• Fixed the nonFatalErrors field not populating in the index node API. (#4615)
• Fixed graph-node panicking on the first startup when both Firehose and RPC providers are configured together. (#4680)
• Fixed block ingestor failing to startup with the error net version for chain mainnet has changed from 0 to 1 when switching from Firehose to an RPC provider. (#4692)
• Fixed Firehose endpoints getting rate-limited due to duplicated providers during connection pool initialization. (#4778)
• Fixed a determinism issue where stale entities where being returned when using get_many and get_derived ([#4801]#4801)

Graphman

• Added two new graphman commands pause and resume: Instead of reassigning to a non-existent node these commands can now be used for pausing and resuming subgraphs. (#4642)
• Added a new graphman command restart to restart a subgraph. (#4742)

Full Changelog: v0.31.0...2891a76

v0.32.0-rc.0

v0.32.0-rc.0 is the first release candidate of the upcoming v0.32.0. This release candidate is intended for testnet use only.

$ docker pull graphprotocol/graph-node:v0.32.0-rc.0

What's New

• Derived fields getter: Derived fields can now be accessed from within the mapping code during indexing. (#4434)
• Sorting interfaces by child entity: Interfaces can now be sorted by non-derived child entities. (#4058)
• File data sources can now be spawned from handlers of other file data sources: This enables the use of file data sources for scenarios where a file data source needs to be spawned from another one. One practical application of this feature is in handling NFT metadata. In such cases, the metadata itself is stored as a file on IPFS and contains embedded IPFS CID for the actual file for the NFT. (#4713)
• Allow redeployment of grafted subgraphs even when graft_base is not available: This will allow renaming of already synced grafted subgraphs even when the graft base is not available, which previously failed due to graft-base validation errors. (#4695)
• history_blocks is now available in the index-node API. (#4662)
• Added a new subgraph features table in primary to easily track information like apiVersion, specVersion, features, and data source kinds used by subgraphs. (#4679)
• subgraphFeatures endpoint now includes data from subgraph_features table.
• ens_name_by_hash is now undeprecated: This reintroduces support for fetching ENS names by their hash, dependent on the availability of the underlying Rainbow Table (#4751).
• Deterministically failed subgraphs now return valid POIs for subsequent blocks after the block at which it failed. (#4774)
• eth-call logs now include block hash and block number: This enables easier debugging of eth-call issues. (#4718)
• Enabled support for substreams on already supported networks. (#4767)
• Add new GraphQL scalar type Int8. This new scalar type allows subgraph developers to represent 8-bit signed integers. (#4511)
• Add support for overriding module params for substreams-based subgraphs when params are provided in the subgraph manifest. (#4759)

Bug fixes

• Fixed PublicProofsOfIndexing returning the error Null value resolved for non-null field proofOfIndexing when fetching POIs for blocks that are not in the cache (#4768)
• Fixed an issue where Block stream would fail when switching back to an RPC-based block ingestor from a Firehose ingestor. (#4790)
• Fixed an issue where derived loaders were not working with entities with Bytes as IDs (#4773)
• Firehose connection test now retries for 30 secs before setting the provider status to Broken (#4754)
• Fixed the nonFatalErrors field not populating in the index node API. (#4615)
• Fixed graph-node panicking on the first startup when both Firehose and RPC providers are configured together. (#4680)
• Fixed block ingestor failing to startup with the error net version for chain mainnet has changed from 0 to 1 when switching from Firehose to an RPC provider. (#4692)
• Fixed Firehose endpoints getting rate-limited due to duplicated providers during connection pool initialization. (#4778)

Graphman

• Added two new graphman commands pause and resume: Instead of reassigning to a non-existent node these commands can now be used for pausing and resuming subgraphs. (#4642)
• Added a new graphman command restart to restart a subgraph. (#4742)

Full Changelog: v0.31.0...c350e4f