Oracle 12.2 RAC : Transaction recovery: lock conflict caught

windsky 2019-01-13

节点一 alert日志:

PDB(17):Transaction recovery: lock conflict caught and ignored

PDB(17):Transaction recovery: lock conflict caught and ignored

PDB(17):Transaction recovery: lock conflict caught and ignored

...

节点二: alert日志

PDB(17):minact-scn: useg scan erroring out with error e:12751

PDB(17):minact-scn: useg scan erroring out with error e:12751

PDB(17):minact-scn: useg scan erroring out with error e:12751

...

起因:

分区表 表级nologging,分批四千万数据并行插入,全表共6亿多数据,导致死锁,查询

select 'alter system kill session '||chr(39)||t2.sid||','||t2.serial#||chr(39)||';'
from v$locked_object t1,v$session t2
where t1.session_id=t2.sid order by t2.logon_time;

alter system kill session '4843,29019';

...

全部是同一个会话,竟有3000多条记录,

后采用杀会话的方式没能释放,又用kill -9 杀掉进程。随后,alert日志出现以上告警。

然后各种查资料...网上有说要dump数据快,技术有限,还好是压力测试用的,等着慢慢恢复,停止此表数据的插入。

 1.查看恢复时使用的回滚段

select b.name useg, b.inst# instid, b.status$ status, a.ktuxeusn
xid_usn, a.ktuxeslt xid_slot, a.ktuxesqn xid_seq, a.ktuxesiz undoblocks,
a.ktuxesta txstatus
from x$ktuxe a, undo$ b
where a.ktuxecfl like '%DEAD%'
and a.ktuxeusn = b.us#;


USEG                  INSTID    STATUS    XID_USN  XID_SLOT    XID_SEQ UNDOBLOCKS TXSTATUS
------------------------------ ---------- ---------- ---------- ---------- ---------- ---------- ----------------
_SYSSMU30_2947991045$            1      3        30    10    50494    3572115 ACTIVE

2.查看恢复进度

select ktuxeusn USN, ktuxeslt Slot, ktuxesqn Seq, ktuxesta State, ktuxesiz Undo from x$ktuxe
  where ktuxesta <> 'INACTIVE'
  and ktuxecfl like '%DEAD%'
  order by ktuxesiz asc;

      USN    SLOT        SEQ STATE            UNDO
---------- ---------- ---------- ---------------- ----------
    30      10      50494 ACTIVE          2815649

等着慢慢恢复,一分钟约一万block的速度,6个小时3572115blocks。第二天观察UNDO恢复正常了,告警也随之消失。

相关推荐

Nexthop / 0评论 2019-10-20