This has happend multiple times now. Whenever our cluster is undergoing GCP maintenance, the 12 sesrvers (its 1 server per node) slowly go down one at a time over the course of an 30 minutes or so and come back on new nodes.
We have a service that is continually issuing queries, ever few seconds and almost without fail, during this maitanence, the aerospike client panics:
This service is a cleanup runner that runs a partition query ever 5 seconds in order to see if any rows match a filter and should be
removed. it is calling "QueryPartitions" in the latest v5.8 client.
Our aerospike cluster version is 5.7.0.8.
This has happend multiple times now. Whenever our cluster is undergoing GCP maintenance, the 12 sesrvers (its 1 server per node) slowly go down one at a time over the course of an 30 minutes or so and come back on new nodes.
We have a service that is continually issuing queries, ever few seconds and almost without fail, during this maitanence, the aerospike client panics:
This service is a cleanup runner that runs a partition query ever 5 seconds in order to see if any rows match a filter and should be removed. it is calling "QueryPartitions" in the latest v5.8 client. Our aerospike cluster version is 5.7.0.8.