Run checks in parallel

The execution and check of the result of the individual
checks were made in a serial stream processing, which means
that the healthcheck endpoint could have actually failed
even if the individual healthchecks would have all completed
in a timely manner.

Introduce the parallel processing of the stream so that
all checks are running in parallel and have more chances
to be completed within the expected timeout from the
load balancer.

Change-Id: Ibbcbbc429bab8c4694f626478a1e7832eac6f64b

src/main/java/com/googlesource/gerrit/plugins/healthcheck/check/GlobalHealthCheck.java

diff --git a/src/main/java/com/googlesource/gerrit/plugins/healthcheck/check/GlobalHealthCheck.java b/src/main/java/com/googlesource/gerrit/plugins/healthcheck/check/GlobalHealthCheck.java
index fa5f14a..d955e47 100644
--- a/src/main/java/com/googlesource/gerrit/plugins/healthcheck/check/GlobalHealthCheck.java
+++ b/src/main/java/com/googlesource/gerrit/plugins/healthcheck/check/GlobalHealthCheck.java
@@ -48,7 +48,7 @@
     Iterable<HealthCheck> iterable = () -> healthChecks.iterator();
     long ts = System.currentTimeMillis();
     Map<String, Object> checkToResults =
-        StreamSupport.stream(iterable.spliterator(), false)
+        StreamSupport.stream(iterable.spliterator(), true)
             .map(check -> Arrays.asList(check.name(), check.run()))
             .collect(Collectors.toMap(k -> (String) k.get(0), v -> v.get(1)));
     long elapsed = System.currentTimeMillis() - ts;