fix(metric-engine): properly propagate errors during batch open operation

[WenyXu]

I hereby agree to the terms of the GreptimeDB CLA.

Refer to a related PR or issue link (optional)

What's changed and what's your intention?

This PR fixes error propagation issues in the metric engine's batch open operation. Previously, underlying errors during batch open operation were not properly propagated to the caller, which could lead to bad user experience.

After:

│ {"timestamp":"2025-06-16T12:36:05.412304Z","level":"INFO","fields":{"message":"going to open 1 region(s)"},"target":"datanode::datanode"}                                        │
│ {"timestamp":"2025-06-16T12:36:05.413249Z","level":"INFO","fields":{"message":"Try to open region 4509732438016(1050, 16777216), worker: 0"},"target":"mito2::worker::handle_ope │
│ n"}                                                                                                                                                                              │
│ {"timestamp":"2025-06-16T12:36:05.413674Z","level":"INFO","fields":{"message":"Try to open region 4509715660800(1050, 0), worker: 0"},"target":"mito2::worker::handle_open"}     │
│ {"timestamp":"2025-06-16T12:36:05.414027Z","level":"INFO","fields":{"message":"Checkpoint not found in data/greptime/public/1050/1050_0000000000/metadata/manifest/, build manif │
│ est from scratch"},"target":"mito2::manifest::manager"}                                                                                                                          │
│ {"timestamp":"2025-06-16T12:36:05.414202Z","level":"INFO","fields":{"message":"Checkpoint not found in data/greptime/public/1050/1050_0000000000/data/manifest/, build manifest  │
│ from scratch"},"target":"mito2::manifest::manager"}                                                                                                                              │
│ {"timestamp":"2025-06-16T12:36:05.415271Z","level":"ERROR","fields":{"message":"Failed to open region: 4509715660800(1050, 0)","err":"0: Failed to open mito region, region type │
│ : metadata, at src/metric-engine/src/engine/open.rs:106:14\n1: Failed to open region, at /home/weny/Projects/greptimedb/src/mito2/src/worker/handle_open.rs:140:54\n2: Incompati │
│ ble WAL provider change. This is typically caused by changing WAL provider in database config file without completely cleaning existing files. Global provider: `raft_engine`, r │
│ egion provider: `kafka`"},"target":"datanode::region_server","span":{"name":"handle_batch_open_requests"},"spans":[{"name":"handle_batch_open_requests"}]}                       │
│ Error: 0: Failed to start datanode, at /home/weny/Projects/greptimedb/src/cmd/src/datanode/builder.rs:128:14                                                                     │
│ 1: Unexpected, violated: Failed to open batch regions: 0: Failed to open mito region, region type: metadata, at src/metric-engine/src/engine/open.rs:106:14                      │
│ 1: Failed to open region, at /home/weny/Projects/greptimedb/src/mito2/src/worker/handle_open.rs:140:54                                                                           │
│ 2: Incompatible WAL provider change. This is typically caused by changing WAL provider in database config file without completely cleaning existing files. Global provider: `raf │
│ t_engine`, region provider: `kafka`, at /home/weny/Projects/greptimedb/src/datanode/src/region_server.rs:802:14

Before:

│ {"timestamp":"2025-06-16T08:55:11.003324Z","level":"INFO","fields":{"message":"Try to open region 4449602895872(1036, 16777216), worker: 0"},"target":"mito2::worker::handle_open"}                    │
│ {"timestamp":"2025-06-16T08:55:11.003817Z","level":"INFO","fields":{"message":"Try to open region 4449586118656(1036, 0), worker: 0"},"target":"mito2::worker::handle_open"}                           │
│ {"timestamp":"2025-06-16T08:55:11.004311Z","level":"INFO","fields":{"message":"Checkpoint not found in data/greptime/public/1036/1036_0000000000/metadata/manifest/, build manifest from scratch"},"ta │
│ rget":"mito2::manifest::manager"}                                                                                                                                                                               │
│ {"timestamp":"2025-06-16T08:55:11.004408Z","level":"INFO","fields":{"message":"Checkpoint not found in data/greptime/public/1036/1036_0000000000/data/manifest/, build manifest from scratch"},"target │
│ ":"mito2::manifest::manager"}                                                                                                                                                                                   │
│ {"timestamp":"2025-06-16T08:55:11.005561Z","level":"ERROR","fields":{"message":"Failed to open batch regions","err":"0: Failed to open batch regions, at /home/weny/Projects/greptimedb/src/datanode/s │
│ rc/region_server.rs:763:14\n1: Physical region 4449586118656(1036, 0) not found, at src/metric-engine/src/engine/open.rs:78:18"},"target":"datanode::region_server","span":{"name":"handle_batch_open_requests" │
│ },"spans":[{"name":"handle_batch_open_requests"}]}                                                                                                                                                              │
│ Error: 0: Failed to start datanode, at /home/weny/Projects/greptimedb/src/cmd/src/datanode/builder.rs:128:14                                                                                           │
│ 1: Unexpected, violated: Failed to open batch regions: 0: Failed to open batch regions, at /home/weny/Projects/greptimedb/src/datanode/src/region_server.rs:763:14                                     │
│ 1: Physical region 4449586118656(1036, 0) not found, at src/metric-engine/src/engine/open.rs:78:18, at /home/weny/Projects/greptimedb/src/datanode/src/region_server.rs:802:14    

PR Checklist

Please convert it to a draft if some of the following conditions are not met.

• I have written the necessary rustdoc comments.
• I have added the necessary unit tests and integration tests.
• This PR requires documentation updates.
• API changes are backward compatible.
• Schema or data changes are backward compatible.

[Copilot]

Pull Request Overview

This PR fixes error propagation issues in the metric engine’s batch open operation by adjusting error handling and consolidating state recovery.

• Introduces a new error variant (NoOpenRegionResult) for missing open region results.
• Refactors error propagation and state recovery flows in the engine’s sync, open, and catchup modules.

Reviewed Changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 1 comment.

File	Description
src/metric-engine/src/error.rs	Adds new error variant and updates error status mapping.
src/metric-engine/src/engine/sync.rs	Removes redundant primary key encoding retrieval in state recovery.
src/metric-engine/src/engine/open.rs	Refactors open logic; converts physical region IDs to a HashMap and adjusts error propagation.
src/metric-engine/src/engine/catchup.rs	Updates state recovery to align with new error propagation design.

Comments suppressed due to low confidence (2)

src/metric-engine/src/engine/open.rs:103

• [nitpick] Consider logging additional context or including a comment explaining the error wrapping with BoxedError to help future debugging of failure cases in open_physical_region_with_results.

let _ = metadata_region_result.context(NoOpenRegionResultSnafu {

src/metric-engine/src/engine/catchup.rs:82

• [nitpick] Since the primary key encoding is now retrieved internally within recover_states, please add an inline comment or update the function’s docstring to document this design change for clarity.

self.recover_states(region_id, physical_region_options)

[fengjiachun]

LGTM