mod_cluster reaches seg fault crashes with unresponsive backend servers

Issue

• We have warns like the following recurring every 5 minutes or so:

[Thu Dec 01 23:51:59.294106 2022] [:warn] [pid 8335:tid 8701] manager_handler STATUS error: MEM: Can't read node with "jvm1" JVMRoute

• Then httpd crashes with a bad URI:

[Thu Dec 01 20:36:57.826665 2022] [proxy:error] [pid 15192:tid 15500] [remote (null):0] AH00898: URI cannot be parsed: ://:0/ returned by /
[Thu Dec 01 20:36:58.648415 2022] [core:notice] [pid 31893:tid 31893] AH00052: child pid 15192 exit signal Segmentation fault (11)

• Or an assertion failure:

[Thu Dec 01 20:31:17.752539 2022] [core:crit] [pid 5163:tid 5464] AH00102: [Thu Dec 01 20:31:17 2022] file mod_proxy_cluster.c, line 786, assertion "!strcmp(worker->s->hostname, helper->shared->hostname)" failed
[Thu Dec 01 20:31:18.190122 2022] [core:notice] [pid 31893:tid 31893] AH00052: child pid 5163 exit signal Aborted (6)