[Chore] Add systemd unit restart policy

Problem: By default, systemd services generated from the NixOS system
configuration don't attempt to restart on failure since
Restart=no. However, in some cases, running processes can fail for
unclear reasons, and the simplest way to bring the failed service back
to life is to restart it. Instead, currently, the service will fail
and trigger an alert without attempting to restart.

Solution: Add default values for startLimitBurst,
startLimitIntervalSec, Restart, and RestartSec.

modules/vault-secrets.nix

@@ -1,11 +1,11 @@
-# SPDX-FileCopyrightText: 2020 Serokell <https://serokell.io/>
+# SPDX-FileCopyrightText: 2020-2023 Serokell <https://serokell.io/>
 #
 # SPDX-License-Identifier: MPL-2.0
 
 { config, lib, pkgs, ... }:
 let
   cfg = config.vault-secrets;
-  inherit (lib) mkMerge flip mapAttrs' nameValuePair optional;
+  inherit (lib) mkMerge mkDefault flip mapAttrs' nameValuePair optional;
 in
 {
   options = import ./options.nix { inherit lib cfg pkgs; };
@@ -26,11 +26,16 @@ in
 
         script = import ./script.nix { inherit cfg scfg lib name; };
 
+        startLimitBurst = mkDefault 5;
+        startLimitIntervalSec = mkDefault 300;
+
         serviceConfig = {
           EnvironmentFile = scfg.environmentFile;
           RemainAfterExit = true;
           Type = "oneshot";
           UMask = "0077";
+          Restart = mkDefault "on-failure";
+          RestartSec = mkDefault 10;
         };
       }
     ))] ++ (flip lib.mapAttrsToList cfg.secrets (