11:29:52,560 ERROR (testng-Test:[]) [TestSuiteProgress] Test failed: org.infinispan.client.hotrod.xsite.SiteDownFailoverTest.testFailoverAfterSiteShutdown
org.infinispan.client.hotrod.exceptions.HotRodClientException: org.infinispan.util.concurrent.TimeoutException: ISPN000427: Timeout after 15 seconds waiting for acks. Id=12

The test starts 2 sites, LON and NYC, shuts down LON to make the client switch to NYC, then starts LON back up and shuts down NYC to make the client switch back to LON.

The primary wants to send back up the put command to NYC, but NYC is down.
If the primary owner is also the site master it works out ok: site master gets a SITE_UNREACHABLE message and cancels the backup request immediately with a {{SuspectException: ISPN000400: Node SiteAddress

was suspected}}.
If the primary owner is not the site master, it doesn't get the SITE_UNREACHABLE message, so the xsite backup request times out only after 15 seconds, just like the triangle request.
Since the originator is the site master, not the primary, it's also waiting for 15s for a triangle backup from the primary, and the triangle backup usually times out first.

From the JGroups trace logs it also seems that non-site masters should receive SITE_UNREACHABLE events.

11:53:42,961 TRACE (jgroups-8,Test-NodeE:[]) [UDP] Test-NodeE: sending msg to Test-NodeF:LON-1, src=Test-NodeE:LON-1, size=250, headers are TEST_RELAY2: SITE_UNREACHABLE [dest=SiteMaster(NYC-2), sender=null, sites=null], UNICAST3: DATA, seqno=1, conn_id=1, first, TP: [cluster=LON-1-org.infinispan.client.hotrod.xsite.SiteDownFailoverTest]