Hello!  We just experienced a failure, where Jenkins would not come up after restart.

```

Oct 16 22:48:49 ip-172-16-53-13 jenkins_logs[10332]: 2023-10-16 22:48:49.072+0000 [id=26]#011SEVERE#011hudson.util.BootFailure#publish: Failed to initialize Jenkins
Oct 16 22:48:49 ip-172-16-53-13 jenkins_logs[10332]: java.lang.InterruptedException
Oct 16 22:48:49 ip-172-16-53-13 jenkins_logs[10332]: at java.base/java.lang.Object.wait(Native Method)
Oct 16 22:48:49 ip-172-16-53-13 jenkins_logs[10332]: at java.base/java.lang.Object.wait(Object.java:328)
Oct 16 22:48:49 ip-172-16-53-13 jenkins_logs[10332]: at org.jvnet.hudson.reactor.Reactor.execute(Reactor.java:288)
Oct 16 22:48:49 ip-172-16-53-13 jenkins_logs[10332]: at jenkins.InitReactorRunner.run(InitReactorRunner.java:49)
Oct 16 22:48:49 ip-172-16-53-13 jenkins_logs[10332]: at jenkins.model.Jenkins.executeReactor(Jenkins.java:1199)
Oct 16 22:48:49 ip-172-16-53-13 jenkins_logs[10332]: at jenkins.model.Jenkins.<init>(Jenkins.java:987)
Oct 16 22:48:49 ip-172-16-53-13 jenkins_logs[10332]: at hudson.model.Hudson.<init>(Hudson.java:86)
Oct 16 22:48:49 ip-172-16-53-13 jenkins_logs[10332]: at hudson.model.Hudson.<init>(Hudson.java:82)
Oct 16 22:48:49 ip-172-16-53-13 jenkins_logs[10332]: at hudson.WebAppMain$3.run(WebAppMain.java:247)
Oct 16 22:48:49 ip-172-16-53-13 jenkins_logs[10332]: Caused: hudson.util.HudsonFailedToLoad
Oct 16 22:48:49 ip-172-16-53-13 jenkins_logs[10332]: at hudson.WebAppMain$3.run(WebAppMain.java:264)
Oct 16 22:48:49 ip-172-16-53-13 jenkins_logs[10332]: 2023-10-16 22:48:49.075+0000 [id=26]#011SEVERE#011h.i.i.InstallUncaughtExceptionHandler$DefaultUncaughtExceptionHandler#uncaughtException: A thread (Jenkins initialization thread/26) died unexpectedly due to an uncaught exception. This may leave your server corrupted and usually indicates a software bug.

```

We are on Jenkins 2.401.3

We were able to get Jenkins to start by renaming the "nodes" folder and creating a new empty "nodes" folder.

I then checked all the config.xml files in the old folder, and they are all well-formed and valid XML.

We did change the slave instance type earlier today, not sure if that had anything to do with it.