前两天看到一份sp报告，在wait events事件中，有：

                                                                        Avg
                                                     Total Wait   wait    Waits
Event                               Waits   Timeouts   Time (s)   (ms)     /txn
---------------------------- ------------ ---------- ---------- ------ --------
db file sequential read         4,490,735          0     14,958      3 3,024.1
control file parallel write         5,256          0        179     34      3.5
db file scattered read             51,799          0        100      2     34.9
class slave wait                       10         10         51   5117      0.0

看到有class slave wait等待avg wait（ms）居然是5117，一直没弄明白为啥能这么高？？

找了几天，最后想起来metalink上search下，居然发现是bug：

Applies to:

Enterprise Manager for RDBMS - Version: 10.1.0.3.0
This problem can occur on any platform.

Symptoms

"class slave wait" is the top wait event on AWR snapshot.
It is grouped under wait class "other"

Output of ADDM reports for these AWR snapshots:

FINDING 1: 100% impact (3431 seconds)
-------------------------------------
Wait class "Other" was consuming significant database time.

   NO RECOMMENDATIONS AVAILABLE

   ADDITIONAL INFORMATION: Database latches in the "Other" wait class were not
      consuming significant database time.

Cause

Bug 3876475

Hdr: 3876475 10.1 RDBMS 10.0 VOS PRODID-5 PORTID-453
Abstract: MIS-CLASSIFICATION OF KSV SLAVE CLASS WAIT

"class slave wait" has the wrong wait classification.It should be marked as an "idle" wait

Solution

Issue is fixed in 10.1.0.4 and 10.2

Upgrade to either 10.1.0.4 or 10.2

数据库版本恰好就是oracle 10.1.0.3.0，这个点背啊！！！

ITPub博客

Subject: "class slave wait" is the top wait event on AWR snapshot

Applies to:

Symptoms

Cause

Solution

最新文章