Forum Discussion
mart_58302
Nimbostratus
NimbostratusiRule debug is set to 0 but still logs are running
Hello.
BIG-IP 9.4.7 Build 320.1 Final, I'm using Aaron -s super Http Https Single Virtual Server iRule - http://devcentral.f5.com/Wiki/default.aspx/iRules/HttpHttpsSingleVirtualServer.html
And debug is disabled:
Set this option to 1 to log debug messages (to /var/log/ltm by default)
set ::debug 0
Still logs are active:
Jun 15 16:11:51 tmm tmm[1852]: Rule http_to_https : ---------------------------- Start ----------------------------
Jun 15 16:11:51 tmm tmm[1852]: Rule http_to_https : HTTPS request from 172.172.31.210 to 192.168.121.138:443
Jun 15 16:11:51 tmm tmm[1852]: Rule http_to_https : ---------------------------- Start ----------------------------
Why?
- The_BhattmanI have no answer, but I would comment out the debug statements and see if that makes a difference. The idea is that whatever is loaded in memory should not have it active it if remove it.
CB 
hoolioCirrostratus
I'd guess it's because there is another iRule which sets a global variable named debug to 1. If you change this iRule from ::debug to ::single_vs_debug, you should be able to avoid the problem.
I've updated the Codeshare example's logging to do this and add the client IP:port to the log statements.
Aaron
mart_58302Strange, I didn't find such another rule, but thanks for the suggestion, I'll try it at tomorrow.
I taught (didn't check syslog before) that maybe 9.4.7 did that, as just upgraded from 9.4.6 to 9.4.7.
BTW, right now I copy&pasted identical iRules for multiple virtual server, but what is the best practice for using one iRule for multiple virtual server?
Edit, reverted to 9.4.6 anyway, as virtual server with client and server certificate inside forwarded about hundred connections after real user's browser one connection and target application went mad.
9.4.6 does work well. Anyone else noticed such bug?- hoolioIf you want to use the exact same iRule for multiple virtual servers you should reference the same iRule. If you need to make changes to the iRule and have different versions of the same iRule saved on the same LTM unit, you should rename all of the variables to be unique to that iRule. It's something of a nuisance, but necessary to avoid trampling of variables between iRules.
If you still see non-iRule issues after upgrading again to 9.4.7, you should open a case with F5 Support.
Aaron - mart_58302Well, after downgrading to 9.4.6 no such running debug logs as in 9.4.7 using same rules.
So seems something is changed in 9.4.7 (however I didn't find any info from changelog) on this is a bug.
I'll try to upgrade only one node to 9.4.7 and change the rule in that node and check the logs. But seems with Aaron -s suggest I can shut down these logs, but I have no idea how disable this flooding issue.
