RAC on virtual machines gsd lsn often UNKNOWN Solution

State online can be started only after the dual-node RAC is successfully installed on the VM. you can always have several unknown machines after each restart, if you can start them manually. crs_start ora. XXXXX, or crs_start-F XXXXX

If no error is reported, it can be started successfully. after searching online, some people say it may be a problem with time synchronization. If it is 10 seconds, there will be no problem. if you want to synchronize time, You can implement a time synchronization mechanism on the virtual machine. when this 10-second gap occurs, starting the system on a virtual machine is slow, and there is always a series of differences when the system is started at the same time. it is estimated that the time-out between two nodes is more than 10 seconds due to the speed of starting resources.

What if one node is started?

Start the machine on node 1 first

[oracle@node1 ~]$ cd $DBSRCIPTHOME[oracle@node1 dbsrcipt]$ sh crs_stat_unknow.sh                   

-- You can't wait to view the status of HA Resource Target State ----------- ------ ora. MYRAC. MYRAC1.inst ONLINE OFFLINEora. MYRAC. MYRAC2.inst ONLINE OFFLINEora. MYRAC. db ONLINE OFFLINEora. node1.ASM1. asm ONLINE on node1ora. node1.LISTENER1 _ NODE1.lsnr ONLINE OFFLINEora. node1.gsd ONLINE OFFLINEora. node1.ons ONLINE on node1ora. node1.vip ONLINE on node1ora. node2.ASM2. asm ONLINE OFFLINEora. node2.LISTENER1 _ NODE2.lsnr ONLINE OFFLINEora. node2.gsd ONLINE OFFLINEora. node2.ons ONLINE OFFLINEora. node2.vip ONLINE on node1

-- The ons vip ASM1 of the result Node 1 is up, while the other ones are starting. You can see that VIP2 has been mounted on node 1.

[oracle@node1 dbsrcipt]$ sh crs_stat_unknow.sh      

-- In a few minutes, we can see that all the services on node 1 are up. GSD and LSNR do not have the UNKNOWS State HA Resource Target State ----------- ------ ----- ora. MYRAC. MYRAC1.inst ONLINE on node1ora. MYRAC. MYRAC2.inst ONLINE OFFLINEora. MYRAC. db ONLINE on node1ora. node1.ASM1. asm ONLINE on node1ora. node1.LISTENER1 _ NODE1.lsnr ONLINE on node1ora. node1.gsd ONLINE on node1ora. node1.ons ONLINE on node1ora. node1.vip ONLINE on node1ora. node2.ASM2. asm ONLINE OFFLINEora. node2.LISTENER1 _ NODE2.lsnr ONLINE OFFLINEora. node2.gsd ONLINE OFFLINEora. node2.ons ONLINE OFFLINEora. node2.vip ONLINE on node1

-- Start Node 2. After several minutes, we can see that node 2 has started three services, other services are starting [Oracle @ node1 dbsrcept] $ sh crs_stat_unknow.shha resource target State ----------- ------ ----- ora. myrac. myrac1.inst online on node1ora. myrac. myrac2.inst online offlineora. myrac. DB online on node1ora. node1.asm1. ASM online on node1ora. node1.listener1 _ node1.lsnr online on node1ora. node1.gsd online on node1ora. node1.ons online on node1ora. node1.vip online on node1ora. node2.asm2. ASM online on node2ora. node2.listener1 _ node2.lsnr online offlineora. node2.gsd online offlineora. node2.ons online on node2ora. node2.vip online on node2 [Oracle @ node1 dbsrcipt] $ sh crs_stat_unknow.sh

-- In a few minutes, the node 2 service is fully started, and no unknowns exist, and VIP2 has been transferred back to node 2, HA resource target State ----------- ------ ----- ora. myrac. myrac1.inst online on node1ora. myrac. myrac2.inst online on node2ora. myrac. DB online on node1ora. node1.asm1. ASM online on node1ora. node1.listener1 _ node1.lsnr online on node1ora. node1.gsd online on node1ora. node1.ons online on node1ora. node1.vip online on node1ora. node2.asm2. ASM online on node2ora. node2.listener1 _ node2.lsnr online on node2ora. node2.gsd online on node2ora. node2.ons online on node2ora. node2.vip online on node2 [Oracle @ node1 dbsrcipt] $