Cloud Tie breaker for GCP

Hi Team,

In our environment all the DB2 db's are running in GCP cloud environment which are enabled with HADR along with TSA. We use network tie breaker in our environment, recently we faced HADR split brain issue due to network issue and raised case with IBM (TS003746592) for the same.

Below are the response from IBM for our issue:
This is a known issue/limitation of the network tiebreaker, RSCT uses UDP for heartbeats, the network tiebreaker uses ICMP, when there is a network problem and the network recovers the ICMP protocol is brought up before UDP is , so the tiebreaker can be reserved while the UDP (RSCT heartbeat) is still failing.
This ends up with a domain_merge
The solution is that:
1) don't use network tiebreaker, use arbitrator node, NFS, Cloud or disk tiebreaker instead
2) don't have network failures
3) increase the period/sensitivity so that longer network outages are not reported by RSCT

many DB2 db's are critical in our environment so we need to have robust solution for this issue. We prefer to change our TSA tie breaker from network to cloud. But for cloud IBM provided solution is for AWS. We need same solution for GCP, We checked the same with google and they told us to inform IBM regarding same. This is the reason i raised this case. We need to your help regarding this solution. Please let us know if any other details required.

https://www.ibm.com/support/knowledgecenter/en/SSRM2X_4.1.0/com.ibm.samp.doc_4.1/concepts_cloud_tiebreaker.html