[Bug]
维护者通常 3 天内回复
评估
调研方向
Start by reproducing with the SQL in the report: create a table, run gpcheckcat on the database, then create the materialized view and run it again. The failing check is inconsistent_pg_class, which reports relmvrefcount as 1 on the coordinator (content -1) and 0 on the segments. Done means gpcheckcat is clean after the materialized view is created. Pull request #2099 is already open against this issue, so review it before starting new work.
由索引模型根据 Issue 内容生成。
描述
Apache Cloudberry version
Apache Cloudberry 2.1.0-incubating build bdf90c5518f
What happened
gpcheckcat is reporting issues on a database with materialized views
Relation oid: 7016964
Relation schema: XXXX
Relation name: XXXXXX
Name of test which found this issue: inconsistent_pg_class
relmvrefcount is '1' on content -1 (XXXXXX:5432)
relmvrefcount is '0' on all other segments
Gpcheckcat is showing no errors before creation of materialized view
After a materialized view is created and run the gpcheckcat is reporting inconsistencies in pg_class
Seems like relmvrefcount is not being updated on the segments
What you think should happen instead
Either relmvrefcount is updated correctly or gpcheckcat should ignore this test. We have multiple CB clusters with multiple databases. And our daily online catalog check is failing on any database with MV
How to reproduce
CREATE TABLE public.test_table (
col1 int,
col2 int
)
DISTRIBUTED randomly;
run gpcheckcat This shows no error
CREATE MATERIALIZED VIEW public.test_table_mv
WITH (appendonly = true, blocksize='32768', compresslevel='3', compresstype=zstd, checksum='true') AS
SELECT * from public.test_table;
run gpcheckcat
Relation oid: 7018655
Relation schema: public
Relation name: test_table
Name of test which found this issue: inconsistent_pg_class
relmvrefcount is '1' on content -1 (XXXXXX:5432)
relmvrefcount is '0' on all other segments
Operating System
NAME="Red Hat Enterprise Linux" VERSION="9.7 (Plow)"
Anything else
Every time. We can reproduce this at will
Are you willing to submit PR?
- Yes, I am willing to submit a PR!
Code of Conduct
- I agree to follow this project's Code of Conduct.
- 主要语言
- C
- 星标
- 1.4k
- 派生
- 260
- 平均合并
- 4 天 5 小时
- 30 天内合并 PR
- 51
环境准备
- 没有 Dockerfile 或 Docker Compose 文件
- 有 Pull Request 模板
- 阅读贡献指南
从这里开始
- 先读完整个 Issue，再读项目的贡献指南。
- 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
- Fork 仓库，在一个分支上完成修改。
- 提交 Pull Request，并在描述里引用这个 Issue 编号。
apache/cloudberry 的其他 Issue
-
[Bug] Expand fails to recognize user-defined tablespaces可能已有人在做 @nix-oss 于 55 天前认领。 未关闭type: Bug
难度 2/5 1-3 小时 新手友好度 76/100
apache/cloudberry#1885 · 2 个 reaction ·
维护者通常 3 天内回复
-
gpexpand first stage fails: 'SyncPackages' object has no attribute 'ret' (extension sync never runs; regression)可能已有人在做 @talmacschen-arch 于 116 天前认领。 未关闭
难度 2/5 1-3 小时 新手友好度 86/100
apache/cloudberry#1825 ·
维护者通常 3 天内回复
-
[Issues] When Initializing a new Cloudberry cluster we are encountering issues可能已有人在做 关联的 PR 仍在进行中或已合并。 未关闭type: Bug
难度 4/5 3-5 天 新手友好度 12/100
apache/cloudberry#2101 · 1 条评论 ·
维护者通常 3 天内回复
-
难度 4/5 3-5 天 新手友好度 12/100
apache/cloudberry#2097 · 1 条评论 ·
维护者通常 3 天内回复
-
type: Bug
难度 4/5 3-5 天 新手友好度 22/100
apache/cloudberry#2091 ·
维护者通常 3 天内回复
查看 apache/cloudberry 的全部 Issue
相似的 Issue
-
docs: AGENTS.md has no rule against running deploy scripts into the owner's game install可能已有人在做 @thesprockee 今天认领。 未关闭
难度 1/5 1 小时以内 新手友好度 85/100
EchoTools/nevr-runtime#524 ·
维护者通常 1 天内回复
-
难度 2/5 1-3 小时 新手友好度 67/100
-
type: feature
难度 1/5 1 小时以内 新手友好度 77/100
-
难度 2/5 1-3 小时 新手友好度 83/100
维护者通常 1 天内回复
-
难度 2/5 1-3 小时 新手友好度 78/100
FujiNetWIFI/fujinet-firmware#1932 ·
维护者通常 1 天内回复