在Docker容器中Oracle11g数据库长时间启动
我使用https://hub.docker.com/r/jaspeen/oracle-11g/ image从转储文件恢复数据库。 提交并从图像运行后，我等待20分钟时打开数据库：
SQL> select status from v$instance; STATUS ------------ MOUNTED 20分钟后
SQL> select status from v$instance; STATUS ------------ OPEN
entrypoint.sh日志：
sqlplus: Database mounted. alertlog: Successful mount of redo thread 1, with mount id 1430168652 alertlog: Database mounted in Exclusive Mode alertlog: Lost write protection disabled alertlog: Completed: ALTER DATABASE MOUNT alertlog: Fri Jan 29 05:49:06 2016 alertlog: ALTER DATABASE OPEN listener: 29-JAN-2016 05:49:06 * service_update * orcl * 0 listener: Fri Jan 29 05:49:29 2016 listener: 29-JAN-2016 05:49:29 * (CONNECT_DATA=(SID=orcl)(CID=(PROGRAM=sqlplus64)(HOST=ubuntu)(USER=user))) * (ADDRESS=(PROTOCOL=tcp)(HOST=10.196.5.92)(PORT=43917)) * establish * orcl * 0 Database mounted in Exclusive Modelistener: Fri Jan 29 05:56:42 2016 listener: 29-JAN-2016 05:56:42 * (CONNECT_DATA=(SID=orcl)(CID=(PROGRAM=sqlplus64)(HOST=ubuntu)(USER=user))) * (ADDRESS=(PROTOCOL=tcp)(HOST=10.196.5.92)(PORT=43925)) * establish * orcl * 0 alertlog: Fri Jan 29 06:04:18 2016 alertlog: Errors in file /opt/oracle/app/diag/rdbms/orcl/orcl/trace/orcl_m000_86.trc: alertlog: ORA-01155: the database is being opened, closed, mounted or dismounted listener: Fri Jan 29 06:06:33 2016 listener: 29-JAN-2016 06:06:33 * (CONNECT_DATA=(SID=orcl)(CID=(PROGRAM=sqlplus64)(HOST=ubuntu)(USER=user))) * (ADDRESS=(PROTOCOL=tcp)(HOST=10.196.5.92)(PORT=43935)) * establish * orcl * 0 listener: Fri Jan 29 06:09:09 2016 listener: 29-JAN-2016 06:09:08 * service_update * orcl * 0 alertlog: Fri Jan 29 06:11:14 2016 alertlog: Thread 1 opened at log sequence 41 alertlog: Current log# 2 seq# 41 mem# 0: /opt/oracle/app/oradata/orcl/redo02.log alertlog: Successful open of redo thread 1 alertlog: Fri Jan 29 06:11:14 2016
为什么这么漫长？
db_recovery_file_dest_size of 3882 MB is 0.00% used. This is a alertlog: user-specified limit on the amount of space that will be used by this alertlog: database for recovery-related files, and does not reflect the amount of alertlog: space available in the underlying filesystem or ASM diskgroup.
可能是每次启动恢复db？ 如何避免它？
我们决定使用iscsi上的zfs作为prod服务器的数据库快照。 在1分钟内启动数据库。