AIX系统月维护查什么（二）

温故：

        在上一篇文章《AIX系统月维护查什么（一）i》中和大家分享了AIX系统如何做月维护（月度巡检），文章主要从系统、磁盘、网络三个方面进行了讲解，但是还不够全。今天给大家补上数据库方面的巡检。

知新：

下面就是需要检查的项目，当然了也可以根据实际情况进行适当裁剪

缓冲区命中率	Statckspack的Buffer Hit	缓冲区的命中率显示相对磁盘读取的数据内存的百分比。
数据字典缓存命中率	select (1-(sum(getmisses)/sum(gets))) * 100 "Hit Ratio" from v$rowcache;	数据字典缓存命中率显示了对数据字典和其它对象的内存读取操作所占的百分比
库缓存命中率	select Sum(Pins)/(Sum(Pins)+Sum(Reloads))* 100 "Hit Ratio" from V$LibraryCache;	库缓存命中率显示了对实际语句和PL/SQL对象的内存读操作所占的百分比
内存排序命中率	select a.value "Disk Sorts", b.value "Memory Sorts",      round((100*b.value)/decode((a.value+b.value),0,1,(a.value+b.value)),2) "Pct Memory Sorts" from   v$sysstat a, v$sysstat b where  a.name ='sorts (disk)' and    b.name ='sorts (memory)';	根据初始化参数PGA_AGGREGATE_TARGET(或者SORT_AREA_SIZE)的值，用户的排序可能在内存中执行或磁盘上的特定的临时表空间执行
最浪费内存的前十个语句占所有语句的比例	select sum(pct_bufgets) from (select rank() over(order by buffer_gets desc) as rank_bufgets,      to_char(100*ratio_to_report(buffer_gets) over(), '999.99')      pct_bufgets        from v$sqlarea) where rank_bufgets < 11;	前十个语句占所有语句的比例过高，将影响Oracle的整体性能，也意味着前十个语句也许需要进行SQL优化
表空间使用情况	select tablespace_name,        total_gb,        free_gb,        used_gb,        used_ratio || '%' used_ratio,        round((used_gb / 90 * 100) - total_gb) "to-90%",        round((used_gb / 85 * 100) - total_gb) "to-85%",        round((used_gb / 80 * 100) - total_gb) "to-80%",        round((used_gb / 75 * 100) - total_gb) "to-75%",        round((used_gb / 70 * 100) - total_gb) "to-70%"   from (select ttt.*           from (select df.tablespace_name,                        df.total_gb,                        nvl(dfs.free_gb, 4) free_gb,                        df.total_gb - nvl(dfs.free_gb, 4) used_GB,                        round((df.total_gb - nvl(dfs.free_gb, 4)) /                              decode(df.total_gb, 0, 1, df.total_gb),                              2) * 100 used_ratio                          from (select tablespace_name,                                round(sum(bytes) / 1024 / 1024 / 1024, 4) total_gb                           from dba_data_files                          group by tablespace_name) df,                        (select tablespace_name,                                round(sum(bytes) / 1024 / 1024 / 1024, 4) free_gb                           from dba_free_space                          group by tablespace_name) dfs                  where df.tablespace_name = dfs.tablespace_name(+)                 union all                 select tt.tablespace_name,                        tt.total_gb,                        round(dt.block_size * ss.fb / 1024 / 1024 / 1024, 4) free_gb,                        tt.total_gb -                        round(dt.block_size * ss.fb / 1024 / 1024 / 1024, 4) used_gb,                        round((dt.block_size * ss.ub / 1024 / 1024 / 1024) /                              decode(tt.total_gb, 0, 1, tt.total_gb),                              2) * 100 used_ratio                   from (select tablespace_name,                                sum(used_blocks) ub,                                sum(free_blocks) fb                           from gv$sort_segment                          group by tablespace_name) ss,                        dba_tablespaces dt,                        (select tablespace_name,                                round(sum(bytes / 1024 / 1024 / 1024), 4) total_gb                           from dba_temp_files                          group by tablespace_name) tt                  where tt.tablespace_name = ss.TABLESPACE_NAME                    and tt.tablespace_name = dt.tablespace_name) ttt          order by 5 desc);	使用率是否在阀值之内
数据文件状态		所有数据文件是否online
监听运行情况	lsnrctl -status
数据库Alert日志报警情况		是否有ORA开头的关键字