master_connection: log why we failed to connect

Previously the exception would've been completely ignored, so all we've seen in the logs was the
successful connection attempt, and then closing the connection after sending 0 bytes.
We were not seeing that Thread.wait_timed_read raised an exception due to a file descriptor having a value >1024.

Signed-off-by: Edwin Török <edwin.torok@cloud.com>

ocaml/database/master_connection.ml

@@ -271,7 +271,7 @@ let do_db_xml_rpc_persistent_with_reopen ~host:_ ~path (req : string) :
           !Db_globs.permanent_master_failure_retry_interval ;
         Thread.delay !Db_globs.permanent_master_failure_retry_interval ;
         !Db_globs.restart_fn ()
-    | e -> (
+    | e ->
         error "Caught %s" (Printexc.to_string e) ;
         (* RPC failed - there's no way we can recover from this so try reopening connection every 2s + backoff delay *)
         ( match !my_connection with
@@ -322,9 +322,7 @@ let do_db_xml_rpc_persistent_with_reopen ~host:_ ~path (req : string) :
           debug "%s: Sleep interrupted, retrying master connection now"
             __FUNCTION__ ;
         update_backoff_delay () ;
-        try open_secure_connection () with _ -> ()
-        (* oh well, maybe nextime... *)
-      )
+        D.log_and_ignore_exn open_secure_connection
   done ;
   !result