[Tracking] CCR (Cross-Cluster Replication): issues, features & bug-fix progress

[Ryan19929]

Background

Doris Cross-Cluster Replication (CCR) keeps a target cluster in sync with a source cluster: full initialization via BACKUP/RESTORE, then continuous incremental sync by pulling binlog commit sequences from the source and applying them on the target through transactional ingest binlog. The open-source syncer lives in selectdb/ccr-syncer; Doris FE/BE provide backup/restore, binlog, and ingest APIs.

This is an umbrella / tracking issue to consolidate CCR-related work — bug fixes, feature requests, and full-sync performance improvements — across apache/doris and selectdb/ccr-syncer, so contributors and maintainers have a single place to see status and pick up work. Related umbrella: #66252. Please comment below to add items or volunteer.

---

Open bugs & compatibility breakage

• [Bug] Table-level CCR silently ignores upstream DROP TABLE selectdb/ccr-syncer#654 — Table-level CCR silently ignores upstream DROP TABLE; task keeps polling with no warning. → fix in feat: auto-recover table sync when upstream table recreated selectdb/ccr-syncer#660 (open, lint CI failing — needs review)
• 表级 CCR 任务在 REPLACE TABLE swap=true 场景下可能错过表 id 切换事件 selectdb/ccr-syncer#677 — Table-level CCR may miss table-id switch after REPLACE TABLE … PROPERTIES('swap'='true'); REPLACE_TABLE binlog tableIds only attaches origTblId. (no Doris PR yet — needs BinlogManager.addReplaceTable fix)
• [Bug] get_binlog_info returns not found for existing binlog files #61063 — get_binlog_info returns not found for existing binlog files after tablet clone; CCR incremental ingest fails. (no PR yet)
• 删除任务后立刻重建同名任务时，任务信息存在被误删的问题 selectdb/ccr-syncer#648 — Deleting a job and immediately recreating one with the same name can corrupt or wipe task metadata. → fixed in improvement: optimize job deletion mechanism selectdb/ccr-syncer#659 (merged 2025-12 — issue still open, pending verification & close)
• [Bug] FE backup staging directories may not be cleaned up #65632 — FE backup staging directories under tmp_dir may not be cleaned up after backup jobs. → fix in [fix](backup) clean up FE backup staging directories #65642 (open, needs review) · also in HYDCP/hy-doris#66, #67 (not yet in apache/doris)

Open feature requests

Storage medium sync

Cross-cluster clusters often differ in SSD/HDD policy; restore and MODIFY PARTITION need explicit medium control on both sides.

• [feat](restore) enhance storage medium control #58897 — Enhance restore storage medium control. (open, needs review — large PR)
• [feature](catalog) support medium allocation mode #63011 — Support medium allocation mode (part 1 of [feat](restore) enhance storage medium control #58897 series). (open, needs review)
• feat: support medium sync policy in CCR tasks selectdb/ccr-syncer#624 — Medium sync policy in CCR tasks (CREATE TABLE retry + MODIFY PARTITION medium sync). (draft, conflicting — blocked on [feat](restore) enhance storage medium control #58897 / [feature](catalog) support medium allocation mode #63011)

Replication-num heterogeneity

• feat: Support custom replication_num for downstream cluster selectdb/ccr-syncer#649 — Custom replication_num for downstream cluster. (merged)
• fix: legacy job compatibility for replication_num selectdb/ccr-syncer#658 — Legacy job compatibility for replication_num. (merged)
• fix: Remove BE caching mechanism and simplify replication number validation selectdb/ccr-syncer#651 — Remove BE caching so restore sees current backend set after scale-out. (merged)
• Allow atomic_restore with different replica num. → HYDCP/hy-doris#51 (not yet in apache/doris)

CCR full-sync optimization

CCR full sync is BACKUP → RESTORE; at scale FE memory, concurrency limits, and staging cleanup become bottlenecks.

• Reduce FE memory usage of backup jobs with massive tablets (Gson DOM serialization, giant String materialization, double journal serialization). → HYDCP/hy-doris#63 (open, needs review — not yet in apache/doris)
• [fix](backup) Strip replica info from backup meta #65321 — Strip replica info from backup meta to shrink backup metadata and reduce FE heap pressure. (draft — split out from the backup FE memory work above; restore recreates replicas anyway, so serialized Replica objects are redundant)
• [improvement](fe) Use streaming Gson adapter for RestoreJob Guava Table fields #64492 — Streaming Gson adapter for RestoreJob Guava Table fields (restore-side memory). (closed without merge — re-submit needed)
• Streaming Gson adapter for RestoreJob Guava Table fields (restore-side memory). → HYDCP/hy-doris#49 (not yet in apache/doris)
• [feature](restore) support concurrent backup/restore #61710 — Support concurrent table-level backup/restore. (open, needs review)
• [fix](restore) Preserve cross-database references when restoring views #59580 — Preserve cross-database references when restoring views. (merged)
• [fix](catalog) Handle incomplete dynamic partition properties #63831 — Handle incomplete dynamic partition properties on restore. (merged)
• Clean backup local job dir after completion. → HYDCP/hy-doris#66 (not yet in apache/doris)
• Centralize backup staging directory cleanup on checkpoint replay and job eviction. → HYDCP/hy-doris#67 (not yet in apache/doris)

Single-replica ingest binlog

• Single-replica ingest binlog for 1-replica target clusters (paired FE/BE + syncer changes). (no upstream PRs yet)

CCR ops & FE connectivity

• fix: JobCollector RPC connection reuse selectdb/ccr-syncer#645 — JobCollector RPC connection reuse. (merged — closes ccr和fe之间的连接数过高 selectdb/ccr-syncer#644)
• fix: job_status API not updating after state change selectdb/ccr-syncer#653 — Fix /job_status not updating after Pause/Resume. (merged)
• Return valid master address for NOT_MASTER errors. → HYDCP/hy-doris#53 (not yet in apache/doris)
• Return master address for snapshot RPCs. → HYDCP/hy-doris#59 (not yet in apache/doris)

Open pull requests awaiting review

PR	Repo	Summary	Closes
selectdb/ccr-syncer#660	ccr-syncer	Auto-recover table sync when upstream table is recreated	selectdb/ccr-syncer#654
selectdb/ccr-syncer#624	ccr-syncer	Medium sync policy in CCR tasks (draft)	—
#65642	doris	Clean up FE backup staging directories	#65632
#65321	doris	Strip replica info from backup meta (draft; part of backup FE memory optimization)	—
HYDCP/hy-doris#63	hy-doris	Reduce FE memory for backup jobs with massive tablets	—
#58897	doris	Enhance restore storage medium control	—
#63011	doris	Support medium allocation mode	—
#61710	doris	Support concurrent backup/restore	—

---

Recently completed (last ~12 months)

ccr-syncer

• fix: avoid nil panic when transaction becomes visible selectdb/ccr-syncer#676 — Fix nil panic when transaction becomes visible.
• Fix: Persist Dest.TableId after partialSync in TableSync mode selectdb/ccr-syncer#672 — Persist Dest.TableId after partialSync in TableSync mode.
• improvement: optimize job deletion mechanism selectdb/ccr-syncer#659 — Optimize job deletion to prevent same-name recreate races. Closes 删除任务后立刻重建同名任务时，任务信息存在被误删的问题 selectdb/ccr-syncer#648 (pending issue close).
• fix: legacy job compatibility for replication_num selectdb/ccr-syncer#658 — Legacy job compatibility for replication_num.
• fix: job_status API not updating after state change selectdb/ccr-syncer#653 — Fix /job_status not updating after Pause/Resume.
• fix: Remove BE caching mechanism and simplify replication number validation selectdb/ccr-syncer#651 — Remove BE caching; fix restore with stale backend info after scale-out.
• feat: Support custom replication_num for downstream cluster selectdb/ccr-syncer#649 — Support custom replication_num for downstream cluster.
• fix: JobCollector RPC connection reuse selectdb/ccr-syncer#645 — JobCollector RPC connection reuse. Closes ccr和fe之间的连接数过高 selectdb/ccr-syncer#644.
• tableRecord add delta rows selectdb/ccr-syncer#638 — tableRecord add delta rows.
• refactor: Handle dummy binlog selectdb/ccr-syncer#486 — Handle dummy binlog.
• fix: table sync should reset the shadow index map selectdb/ccr-syncer#662 (w41ter) — Fix INSERT OVERWRITE in TableSync mode. Closes CCR 3.0.6 执行 INSERT OVERWRITE 同步失败 selectdb/ccr-syncer#664.

apache/doris

• [fix](restore) Preserve cross-database references when restoring views #59580 — Preserve cross-database references when restoring views.
• [fix](catalog) Handle incomplete dynamic partition properties #63831 — Handle incomplete dynamic partition properties on restore.

Earlier / closed without merge (for reference)

• [improvement](binlog) detect dropped tables early in binlog API #59044 — Detect dropped tables early in binlog API (stale-closed Jun 2026; referenced by feat: auto-recover table sync when upstream table recreated selectdb/ccr-syncer#660)
• [improvement](fe) Use streaming Gson adapter for RestoreJob Guava Table fields #64492 — RestoreJob Gson streaming (closed without merge — see HYDCP/hy-doris#49)
• [feat](restore): enhanced ccr storage_medium sync #52391 / [enhancement](catalog) add new table property to control storage medium select behavior #53120 — Earlier storage-medium attempts, superseded by [feat](restore) enhance storage medium control #58897 / [feature](catalog) support medium allocation mode #63011
• atomic_restore副本数不同导致失败 selectdb/ccr-syncer#647 — atomic_restore replica mismatch discussion → doris#42337 / HYDCP/hy-doris#51
• 跨版本使用ccr可行吗？从2.1.x->3.0 selectdb/ccr-syncer#652 / version info is incompatable for doris #374 — Cross-version feasibility & getBinlog Q&A (closed)