shaba/lvm2
1
0
Fork 0
mirror of git://sourceware.org/git/lvm2.git synced 2025-01-02 01:18:26 +03:00
Code Releases Activity

tests: better handling for dmeventd restart

Browse Source 
Handle the case, where we  'kill -9' running dmeventd,
and restarting 'dmeventd -R' manages to still contact this
'zombie' damone and manages to get list of monitored devices
out of this and eventually 'run' new and in this case
unexpected instance of dmeventd.
This commit is contained in:
Zdenek Kabelac 2024-04-11 23:43:27 +02:00
parent aae7c98aaa
commit 8ce8e33a44
1 changed files with 26 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

38
test/shell/dmeventd-restart.sh
View File

@ -20,21 +20,22 @@ _restart_dmeventd() {
#rm -f debug.log*
dmeventd -R -fldddd -e "$PWD/test_nologin" > debug.log_DMEVENTD_$RANDOM 2>&1 &
echo $! >LOCAL_DMEVENTD
local local=$!
echo "$local" >LOCAL_DMEVENTD
for i in {1..10}; do
pid=$(pgrep -o dmeventd || true)
test -n "$pid" || break; # no dmeventd running
test "$pid" = "$(< LOCAL_DMEVENTD)" && break
for i in {1..50}; do
pid=$(pgrep -o dmeventd) || break
# Check pid and dmeventd readiness to communicate
test "$pid" = "$local" && dmeventd -i && break
sleep .2
done
# wait a bit, so we talk to the new dmeventd later
if test -n "$pid" ; then
for i in {1..10}; do
dmeventd -i && break;
sleep .1
done
if [ "$i" -eq 50 ]; then
# Unexpected - we waited over 10 seconds and
# dmeventd has not managed to restart
cat /run/dmeventd.pid || true
pgrep dmeventd || true
die "dmeventd restart is too slow"
fi
}
@ -61,7 +62,20 @@ lvchange --monitor y --verbose $vg/4way 2>&1 | tee lvchange.out
test -e LOCAL_CLVMD || grep 'already monitored' lvchange.out
# now try what happens if no dmeventd is running
kill -9 "$(< LOCAL_DMEVENTD)"
pid=$(< LOCAL_DMEVENTD)
kill -9 "$pid"
# TODO/FIXME: it would be surely better, if the wait loop bellow would
# not be need however ATM the API for communication is not welldetecting
# this highly unusual race case - and things will simply timeout on
# reading failure within 4 seconds.
# Fixing would require to add some handling for losing FIFO connection
for i in {1..10}; do
# wait here for a while until dmeventd dies....
# suprisingly it's not instant and we can actually
# obtain list of monitored devices...
test -z $(ps -p "$pid" -o comm=) && break
sleep .1
done
rm LOCAL_DMEVENTD debug.log*
_restart_dmeventd