【测试类型：SQL功能】【测试版本：5.0.1】【概率失败问题】执行建表建索引、删除表操作后，重启集群，备机均无法start，pg日志报错“PANIC: The Page's LSN[15026772743224134984] bigger than want set LSN [353026928]”

【标题描述】: 执行建表建索引、删除表操作后，重启集群，备机均无法start，pg日志报错“PANIC:  The Page's LSN[15026772743224134984] bigger than want set LSN [353026928]”
【测试类型：SQL功能】【测试版本：5.0.1】【概率失败问题】设置一些同步参数后，执行建表建索引及删除表操作后，重启集群，备机无法start，pg日志报错“PANIC:  The Page's LSN[15026772743224134984] bigger than want set LSN [353026928]”

【操作系统和硬件信息】(查询命令: cat /etc/system-release, uname -a):
CentOS Linux release 7.6.1810 (Core)

【测试环境】(单机/1主x备x级联备):
一主两备

【被测功能】:
建表建索引后，重启数据库操作

【测试类型】:
功能测试

【数据库版本】(查询命令: gaussdb -V):
gsql (openGauss 5.0.1 build 91c814e3) compiled at 2023-10-19 10:07:03 commit 0 last mr

【预置条件】:
数据库功能正常

【操作步骤】(请填写详细的操作步骤):
1. 设置guc参数
gs_guc set -N all -D /openGauss/peilq_centos1/cluster/dn1 -c "synchronous_standby_names='*'" &&
gs_guc set -N all -D /openGauss/peilq_centos1/cluster/dn1 -c "synchronous_commit='remote_apply'" &&
gs_guc set -N all -D /openGauss/peilq_centos1/cluster/dn1 -c "hot_standby=on" &&
gs_guc set -N all -D /openGauss/peilq_centos1/cluster/dn1 -c "wal_level='hot_standby'" &&
gs_guc set -N all -D /openGauss/peilq_centos1/cluster/dn1 -c "recovery_max_workers=4" &&
gs_om -t restart

2. 重复执行建表建索引、删除表操作

```
create table t_hash_index_0003 (id int, num int, sex varchar(20) default 'male');
insert into t_hash_index_0003 select random()*10, random()*3, 'XXX' from generate_series(1,5000);
drop index if exists i_hash_index_0003;
create index i_hash_index_0003 on t_hash_index_0003 using hash (id);
explain select count(*) from t_hash_index_0003 where id=100;

gs_om -t restart

create table t_hash_index_0003 (id int, num int, sex varchar(20) default 'male');
insert into t_hash_index_0003 select random()*10, random()*3, 'XXX' from generate_series(1,5000);
drop index if exists i_hash_index_0003;
create index i_hash_index_0003 on t_hash_index_0003 using hash (id);
explain select count(*) from t_hash_index_0003 where id=100;
drop table if exists t_hash_index_0003;
```

3. 重启数据库
gs_om -t restart

【预期输出】:
1. 设置guc参数成功
2. 重复执行建表建索引、删除表操作，有些业务报错（表存在重复报错）
3. 重启数据库成功

【实际输出】:
1. 设置guc参数成功
2. 重复执行建表建索引、删除表操作，有些业务报错（表存在重复报错）
3. 重启数据库失败

【原因分析】:

【日志信息】(请附上日志文件、截图、coredump信息):
备机无法start pg日志
```
2023-10-20 14:57:11.760 653224c7.6131 [unknown] 139743546832640 dn_6002 0 dn_6001_6002_6003 55000  0 [BACKEND] CONTEXT:  xlog redo new_bucket 22, meta_page_masks_updated F, issplitpoint_changed F
2023-10-20 14:57:11.760 653224c7.6131 [unknown] 139743546832640 dn_6002 0 dn_6001_6002_6003 55000  0 [BACKEND] LOG:  page 18 (lsn: 150aa7f8) of relation base/15652/16508 (0/4294967295) is uninitialized
2023-10-20 14:57:11.760 653224c7.6131 [unknown] 139743546832640 dn_6002 0 dn_6001_6002_6003 55000  0 [BACKEND] CONTEXT:  xlog redo clear_dead_marking F, is_primary F
2023-10-20 14:57:11.763 653224c7.6131 [unknown] 139743546832640 dn_6002 0 dn_6001_6002_6003 42809  0 [BACKEND] PANIC:  The Page's LSN[15026772743224134984] bigger than want set LSN [353026928]
2023-10-20 14:57:11.763 653224c7.6131 [unknown] 139743546832640 dn_6002 0 dn_6001_6002_6003 42809  0 [BACKEND] CONTEXT:  xlog redo prevblkno 18, nextblkno 4294967295, ntups 99, is_primary F
2023-10-20 14:57:11.763 653224c7.6131 [unknown] 139743546832640 dn_6002 0 dn_6001_6002_6003 42809  0 [BACKEND] BACKTRACELOG:  tid[5998]'s backtrace:
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(+0xe0c04e) [0x558fa214104e]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(_Z9errfinishiz+0x411) [0x558fa2139281]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(_Z11elog_finishiPKcz+0x1f3) [0x558fa2139613]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(+0x1879e45) [0x558fa2baee45]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(_Z9hash_redoP15XLogReaderState+0x337) [0x558fa2baf1d7]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(_Z15ApplyRedoRecordP15XLogReaderState+0x75) [0x558fa2b28ab5]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(+0x17a41a7) [0x558fa2ad91a7]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(_ZN17parallel_recovery13ApplyProcHeadEPNS_8RedoItemE+0xb94) [0x558fa2ada8b4]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(_ZN17parallel_recovery18PageRedoWorkerMainEv+0x2e0) [0x558fa2adb0a0]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(_Z26GaussDbAuxiliaryThreadMainIL15knl_thread_role22EEiP14knl_thread_arg+0x120) [0x558fa26ba9e0]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(_Z17GaussDbThreadMainIL15knl_thread_role22EEiP14knl_thread_arg+0x2eb) [0x558fa26baceb]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(+0x1358e95) [0x558fa268de95]
        /usr/lib64/libpthread.so.0(+0x7dd5) [0x7f1a09b3fdd5]
        /usr/lib64/libc.so.6(clone+0x6d) [0x7f1a09868ead]
        Use addr2line to get pretty function name and line

2023-10-20 14:57:11.763 653224c7.6131 [unknown] 139743546832640 dn_6002 0 dn_6001_6002_6003 42809  0 [BACKEND] PANIC:  The Page's LSN[15026772743224134984] bigger than want set LSN [353026928]
2023-10-20 14:57:11.763 653224c7.6131 [unknown] 139743546832640 dn_6002 0 dn_6001_6002_6003 42809  0 [BACKEND] CONTEXT:  xlog redo prevblkno 18, nextblkno 4294967295, ntups 99, is_primary F
2023-10-20 14:57:11.763 653224c7.6131 [unknown] 139743546832640 dn_6002 0 dn_6001_6002_6003 42809  0 [BACKEND] BACKTRACELOG:  tid[5998]'s backtrace:
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(+0xe0c04e) [0x558fa214104e]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(_Z9errfinishiz+0x411) [0x558fa2139281]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(_Z11elog_finishiPKcz+0x1f3) [0x558fa2139613]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(+0x1879e45) [0x558fa2baee45]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(_Z9hash_redoP15XLogReaderState+0x337) [0x558fa2baf1d7]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(_Z15ApplyRedoRecordP15XLogReaderState+0x75) [0x558fa2b28ab5]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(+0x17a41a7) [0x558fa2ad91a7]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(_ZN17parallel_recovery13ApplyProcHeadEPNS_8RedoItemE+0xb94) [0x558fa2ada8b4]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(_ZN17parallel_recovery18PageRedoWorkerMainEv+0x2e0) [0x558fa2adb0a0]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(_Z26GaussDbAuxiliaryThreadMainIL15knl_thread_role22EEiP14knl_thread_arg+0x120) [0x558fa26ba9e0]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(_Z17GaussDbThreadMainIL15knl_thread_role22EEiP14knl_thread_arg+0x2eb) [0x558fa26baceb]
        /openGauss/peilq_centos1/cluster/app/bin/gaussdb(+0x1358e95) [0x558fa268de95]
        /usr/lib64/libpthread.so.0(+0x7dd5) [0x7f1a09b3fdd5]
        /usr/lib64/libc.so.6(clone+0x6d) [0x7f1a09868ead]
        Use addr2line to get pretty function name and line

2023-10-20 14:57:19.361 653224c6.1 [unknown] 139750176936256 [unknown] 0 dn_6001_6002_6003 01000  0 [BACKEND] WARNING:  could not fork new process for connection due to PMstate PM_RECOVERY
2023-10-20 14:57:19.361 653224c6.1 [unknown] 139750176936256 [unknown] 0 dn_6001_6002_6003 00000  0 [BACKEND] LOG:  not ready to start snapshot capturer.
```

【测试代码】:

GVP openGauss/openGauss-server

内容风险标识

评论 (7)

GVPopenGauss/openGauss-server

内容风险标识