You need to resize /dev/shm.
And to do that you need to stop database (and clusterware) using it (/dev/shm needs to be empty so you can umount it for resize) . So stop one node database instance (PROD) and then resize /dev/shm on that node.
And then start that instance and make same for the second instance. This way you can resize /dev/shm without taking whole database down.
You can find guide for resize for example here:
previously.. we used to bring up the database with 2gb memory target and that time only 7gb was free space available at tmpfs /dev/shm..
let me know one thing ..
by disabling the AMM and enabling the ASMM will override these ORA-00845..
I checked the one environment where I have ASMM on and it is using /dev/shm .
But there is not same kind of memory requirements as with AMM.
soo... know what would be the solution except mounting the tmpfs with more space..I can't afford to do these because current already another production is running on the same RAC nodes...
change the /dev/shm size on fly:
#mount -t tmpfs tmpfs -o size=12g,remount /dev/shm
it should not interrupt your oracle instance(test it yourself before use on production). it works
Does it work if already an another Prod RAC instance is running....
took you 4days to reply????....
i have showed you the syntax to change your /dev/shm size! increase it to the size that suit your environment.
this is not an test to plan suddenly and follow ur instructions.. This is prod db..and it may take morethan 4days to reply.
and Thanks but i know how to increase the size of /dev/shm
paste the error you are getting....
Hi Preethi DBA
I know things can be hectic with production support and all
But I will recommend creating a service request with oracle if you are not clear on any issue resolution that impacts production ..
Have a great day.