参考文档：

GCS_SERVER_PROCESSES (oracle.com)

在awr报告中，看到addm建议如下：

其中有Global Cache Messaging和 Global Cache Congestion

Activity During the Analysis Period
-----------------------------------
Total database time was 2724 seconds.
The average number of active sessions was .75.Summary of Findings
-------------------Description              Active Sessions      RecommendationsPercent of Activity-----------------------  -------------------  ---------------
1  Top SQL Statements       .6 | 80.47           2
2  Hard Parse               .03 | 3.33           0
3  Global Cache Messaging   .02 | 3.15           1
4  Global Cache Congestion  .02 | 2.4            1

关于Global Cache Congestion的建议如下：

ADDM建议增加参数gcs_server_processes的值 

Finding 4: Global Cache Congestion
Impact is .02 active sessions, 2.4% of total activity.
------------------------------------------------------
Global Cache Service Processes (LMSn) in other instances were not processing
requests fast enough.Recommendation 1: Database ConfigurationEstimated benefit is .02 active sessions, 2.4% of total activity.-----------------------------------------------------------------ActionIncrease throughput of the Global Cache Service (LMSn) processes.Increase the number of Global Cache Service processes by increasing thevalue of the parameter "gcs_server_processes". Alternatively, if thehost is CPU bound consider increasing the OS priority of the GlobalCache Service processes.RationaleThe value of parameter "gcs_server_processes" was "8" during theanalysis period.Symptoms That Led to the Finding:---------------------------------Inter-instance messaging was consuming significant database time on thisinstance.Impact is .02 active sessions, 3.15% of total activity.Wait class "Cluster" was consuming significant database time.Impact is .03 active sessions, 3.33% of total activity.

查看gcs_server_processes参数的设置及相关参数（CPU、内存、SGA）

CPU为96，内存为256G ，SGA为80G 

该RAC中的LMS数量为8 

根据官网对该参数的介绍：

Oracle calculates the default value as follows (in order of precedence):

1. If CLUSTER_DATABASE is set to false, then 0

2. If Oracle ASM, then 1

3. If 1 - 3 CPUS, then 1

4. If 4 - 15 CPUs, then 2

5. If 128 or more CPUs and SGA is 100 GB or more, then (CPUs / 6). If the result includes a fraction, then the fraction is disregarded.

6. Otherwise, the value is 2 + (CPUs / 32). If the result includes a fraction, then the fraction is disregarded. For example, if you have 20 CPUs, then 2 + (20 / 32) would equal 2 GCS processes.

7. On NUMA-enabled systems with 32 or more CPUs, the value is rounded up to a multiple of the number of NUMA processor groups, with a limit of (CPUs / 4) rounded down to a multiple of the number of NUMA processor groups.

　再看一个真实的例子　：　128 CPU，256G内存，SGA 122G

RAC中LMS为21：

根据官方的计算，第五条，128/6 = 21.33333 .去掉小数，21 。 与RAC统计中的lms吻合。 

例子2 : 56 CPU ,256G 内存 ，SGA 52G 

RAC中统计信息中，lms为 2 

根据第六条的计算。 2+（56/32）= 3.75 . RAC中统计LMS=2 。吻合。

GCS_SERVER_PROCESS包含LMS和LM进程。 （上面计算出来gcs_server_process = 3, LMS=2 ）。CPU数量也可能要以数据库的参数cpu_count为准？？ 

GCS_SERVER_PROCESSES specifies the number of background GCS server processes (LMSn and LMnn) to serve the inter-instance traffic among Oracle RAC instances.

END 。