ComputerSurge/bind9
3
0
Fork 0
Code Pull Requests 1 Activity

fixed a bug in rolling timestamp logfiles

Browse Source 
due to comparing logfile suffixes as 32 bit rather than 64 bit
integers, logfiles with timestamp suffixes that should have been
removed when rolling could be left in place. this has been fixed.
This commit is contained in:
Evan Hunt
2021-10-12 16:31:47 -07:00
parent 76baed3343
commit 9a9e906306
3 changed files with 29 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
bin/tests/system/logfileconfig/tests.sh
View File

@@ -187,10 +187,22 @@ ret=0
copy_setports ns1/named.tsconf.in ns1/named.conf
# a seconds since epoch version number
touch ns1/named_ts.1480039317
# a timestamp version number
touch ns1/named_ts.20150101120000120
rndc_reconfig ns1 10.53.0.1 > rndc.out.test$n
$DIG version.bind txt ch @10.53.0.1 -p ${PORT} > dig.out.test$n
grep "status: NOERROR" dig.out.test$n > /dev/null || ret=1
test_with_retry -f ns1/named_ts.1480039317 && ret=1
_found2() (
$DIG version.bind txt ch @10.53.0.1 -p ${PORT} > dig.out.test$n
grep "status: NOERROR" dig.out.test$n > /dev/null || ret=1
# we are configured to keep three versions, so the oldest
# timestamped versions should be gone, and there should
# be two new ones.
[ -f ns1/named_ts.1480039317 ] && return 1
[ -f ns1/named_ts.20150101120000120 ] && return 1
set -- ns1/named_ts.*
[ "$#" -eq 2 ] || return 1
)
retry_quiet 5 _found2 || ret=1
if [ "$ret" -ne 0 ]; then echo_i "failed"; fi
status=$((status+ret))