[Upgrade][GSI] Query returns with error "index_not_found" after offline upgrade of a cluster

Steps:
1. Cluster initialised with build 4.1.0-5005 as given in environment
2. Default and standard_bucket_0 buckets are created.
3. An index is created on each bucket.
4. Couchbase service is stopped on all nodes.
5. Nodes are upgraded to 4.5.0-1301
6. When cluster is back, query the indexes.

cbq> EXPLAIN SELECT job_mo FROM default;
{
"requestID": "1aa563ae-81fa-48ed-b4ae-82ecd33fc4a2",
"signature": "json",
"results": [
{
"#operator": "Sequence",
"~children": [

{ "#operator": "PrimaryScan", "index": "#primary", "keyspace": "default", "namespace": "default", "using": "view" }

,
{
"#operator": "Parallel",
"~child": {
"#operator": "Sequence",
"~children": [

{ "#operator": "Fetch", "keyspace": "default", "namespace": "default" }

,
{
"#operator": "InitialProject",
"result_terms": [

{ "expr": "(`default`.`job_mo`)" }

]
},

{ "#operator": "FinalProject" }

]
}
}
]
}
],
"status": "success",
"metrics":

{ "elapsedTime": "2.95303ms", "executionTime": "2.717758ms", "resultCount": 1, "resultSize": 1287 }

}

No index is shown in the query.
Index stats are not available. In UI, index definitions are visible.
Indexer node is online.
Warning message is shown "Cannot communicate with indexer process.Information on indexes may be stale. Will retry."

Screenshots attached.