Increase timeout in container test

When the container healthcheck was added, it also replaced looping
over trino-cli in the container test script with docker inspect.
Since executing trino-cli could time out after the container started
but before the server was ready, it actually added to the total timeout
of the loop. Now, none of the commands in the loop are blocking for
longer periods, so the timeout is easier to control.

It was measured that the container can take up to 7 minutes to start on arm64
using emulation, so the timeout is raised to 10 minutes.

core/docker/container-test.sh

@@ -8,7 +8,7 @@ function cleanup {
 
 function test_trino_starts {
     local QUERY_PERIOD=10
-    local QUERY_RETRIES=30
+    local QUERY_RETRIES=60
 
     CONTAINER_ID=
     trap cleanup EXIT