Unclean shutdown in WebDAV door if currently killing mover

[paulmillar]

trafficstars

A domain that runs a WebDAV door logged that it failed to shutdown cleanly. The problematic threads had the following stack-trace (as called by Milton)

16 Dec 2020 10:11:09 (WebDAV-uct2-webdav) [] Thread: jetty-29417 [A--] (5) TIMED_WAITING
16 Dec 2020 10:11:09 (WebDAV-uct2-webdav) []     java.lang.Object.wait(Native Method)
16 Dec 2020 10:11:09 (WebDAV-uct2-webdav) []     org.dcache.util.Transfer.waitForMover(Transfer.java:675)
16 Dec 2020 10:11:09 (WebDAV-uct2-webdav) []     org.dcache.util.Transfer.killMover(Transfer.java:1141)
16 Dec 2020 10:11:09 (WebDAV-uct2-webdav) []     org.dcache.util.Transfer.killMover(Transfer.java:1109)
16 Dec 2020 10:11:09 (WebDAV-uct2-webdav) []     org.dcache.webdav.DcacheResourceFactory$HttpTransfer.killMover(DcacheResourceFactory.java:1545)
16 Dec 2020 10:11:09 (WebDAV-uct2-webdav) []     org.dcache.webdav.DcacheResourceFactory.beginRead(DcacheResourceFactory.java:1173)
16 Dec 2020 10:11:09 (WebDAV-uct2-webdav) []     org.dcache.webdav.DcacheResourceFactory.getReadUrl(DcacheResourceFactory.java:1118)
16 Dec 2020 10:11:09 (WebDAV-uct2-webdav) []     org.dcache.webdav.DcacheFileResource.checkRedirect(DcacheFileResource.java:174)

The fall-back process, where the thread is interrupted, was successful, as logged with messages like:

Forcefully interrupting thread jetty-29417 during shutdown