The cause of this story is that there is a classmate in the version 10.2.0.1 (said 9i may also encounter) a host running 198/248/249 (24.9) days after the OCI client spin spin consumption of a large number of CPUs Bug,spin the cause is sltrgatime64 ( function to a dead loop call to The Times () function, and the bug number has "4612267 OCI client spins when machine uptime >= days", "249 client is in a OCI LO OP when MACHINE UPTIME HITS 248 days ".
This bug can let everyone remember, I am afraid it will be because of the number of days and the host is not unrelated to the characteristics of the trigger; 10.2.0.1 is the base release of 10GR2, but also because there are a large number of enterprises in the country to the database = 249 days, "OCI CLIENT is in the" I Nfinite LOOP when MACHINE UPTIME HITS 248 days ".
This bug can be remembered, I am afraid it will be because of the number of days and the host is not unrelated to the characteristics of the trigger; 10.2.0.1 is the base release of 10GR2, but also because a large number of enterprises in the country to the version of the database patch set upgrade not enough attention, so the bug in 07 , from time to time before 08 to the industry's friends to bring trouble.
However, after the bug was discovered, Oracle immediately released the one-off patch on 10.2.0.1 to fix the problem, and also introduced a fix to the bug in subsequent 10.2.0.2 patch set. In other words, unless you are still using version 10.2.0.1, you have no need to worry that the host does not reboot to a certain date will cause Oracle to fail.
Conclusion: Oracle on a host running for more than 198 days may encounter bugs that cause CPU depletion this legend is good for version 10.2.0.1, so it is not possible to say that this information is incorrect. However, there is no need to worry about patch set 10.2.0.2 later versions.
This article is transferred from CUUG official website technical article.
See more highlights of this column: http://www.bianceng.cnhttp://www.bianceng.cn/database/Oracle/