Bug 241581
| Summary: | RHEL4U5 GA : snmpwalk takes 10 to 12 mins to traverse mib-2 tree | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Product: | Red Hat Enterprise Linux 4 | Reporter: | Narendra K <narendra_k> | ||||||||||
| Component: | net-snmp | Assignee: | Jan Safranek <jsafrane> | ||||||||||
| Status: | CLOSED ERRATA | QA Contact: | |||||||||||
| Severity: | high | Docs Contact: | |||||||||||
| Priority: | urgent | ||||||||||||
| Version: | 4.6 | CC: | jfeeney, palliv, pb, rainer.traut, rvokal, wwlinuxengineering | ||||||||||
| Target Milestone: | --- | Keywords: | ZStream | ||||||||||
| Target Release: | --- | ||||||||||||
| Hardware: | i386 | ||||||||||||
| OS: | Linux | ||||||||||||
| Whiteboard: | |||||||||||||
| Fixed In Version: | RHBA-2007-0738 | Doc Type: | Bug Fix | ||||||||||
| Doc Text: | Story Points: | --- | |||||||||||
| Clone Of: | Environment: | ||||||||||||
| Last Closed: | 2007-11-15 16:00:33 UTC | Type: | --- | ||||||||||
| Regression: | --- | Mount Type: | --- | ||||||||||
| Documentation: | --- | CRM: | |||||||||||
| Verified Versions: | Category: | --- | |||||||||||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||||||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||||||||
| Embargoed: | |||||||||||||
| Bug Depends On: | |||||||||||||
| Bug Blocks: | 246028, 251132 | ||||||||||||
| Attachments: |
|
||||||||||||
Created attachment 155534 [details]
The snmpd.conf file used to configure snmpd agent
Created attachment 155535 [details]
the snmpd log /var/log/snmpd.log
Created attachment 155536 [details]
the /var/log/messages log file
This is already fixed upstream - the second log line is not produced in newer versions of snmpd -> syslogd cuts the first log line to 'last message repeated 533 times'. I can port it back to net-snmp-5.1.2. Is there an option to disable this logging at all? I tried removing "-a" from default options - won't help. I played around with the "-L" option, but it didn't help also. As I understand (according to http://www.net-snmp.org/docs/man/snmpcmd.html) I should use for warning priority to syslog: -LSw -Lsd But -LSw isn't supported neither on RHEL4U5 nor in Fedora Core 6 version. Digging through source code I have found, that -LS isn't supported at all at least in 5.3.1 (see snmp_logging.c and look for call of function decode_priority). A short test shows me that "-LFw" works (like expected after digging through source code). Looks like programmers forgot to add/copy this piece of code, but author of man page didn't realize it and one also forgot to test described options. Should I file another bug according to this issue. Ooops, I must revert my comment about the missing function call, it exists on case "S", but then the problem must be somewhere else. -LFn works, -LSn isn't supported - now still don't know why. Digging through modified source code I finally found how the option works: OPTIONS="-LS notice daemon -Lf /dev/null -p /var/run/snmpd.pid" This will suppress all the messages mentioned in #1 Looks like one have at least to update the man page for a proper working example. This request was evaluated by Red Hat Product Management for inclusion in a Red Hat Enterprise Linux maintenance release. Product Management has requested further review of this request by Red Hat Engineering, for potential inclusion in a Red Hat Enterprise Linux Update release for currently deployed products. This request is not yet committed for inclusion in an Update release. With respect to comment #4 could you make the patch available so that i can test it. Created attachment 156962 [details]
patch for this bug
The patch is attached. You can find .srpm with it + other patches for RHEL4.6 at http://people.redhat.com/jsafrane/bugs/241581/ With respect to comment #14, tested the application built using the srpm mentioned in comment #14. The messsages mentioned in the description of the problem are not repeatedly being logged into /var/log/messages. And snmpwalk on mib-2 tree takes about 10 seconds. *** Bug 246057 has been marked as a duplicate of this bug. *** HP tested and passed verification. people.page lists net-snmp-5.1.2-11.EL4.10.test.src.rpm in comment #14. RHEL4.6-snapshot4 contains net-snmp-5.1.2-11.EL4.11.i386.rpm which should contain the fix. An advisory has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on the solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHBA-2007-0738.html |
Description of problem: When the mib-2 tree is walked on RHEL4 U5 GA, it takes anywhere between 7 to 12 mins, with the default "-Lsd -Lf /dev/null - p /var/run/snmpd.pid -a" option to snmpd. If "-Lsd" option is not supplied it takes 10 to 12 seconds to walk the tree. The following messages are found many times in /var/log/messages. snmpd[15029]: Connection from - 172.0.0.1 snmpd[15029]: transport socket = 12 Version-Release number of selected component (if applicable): net-snmp-5.1.2-11.EL4.10 kernel version : 2.6.9-55.ELsmp How reproducible: Often Steps to Reproduce: 1.Install RHEL4 Update 5 GA. 2.Add the following line to /etc/snmp/snmpd.conf: view all included .1 3.Change the "access" line in /etc/snmp/snmpd.conf to: access notConfigGroup "" any noauth exact all none none 4. Run "service snmpd start" to start snmpd. 5. Run "date; snmpwalk -v1 -c public localhost mib-2 >snmpwalk.log; date" to walk MIB-II and timestamp the start and end of the walk. Actual results: It takes about 7 to 12 mins to walk the tree Expected results: It should take about 10 to 12 seconds. Additional info : If *.info string is removed from /etc/syslog.conf snmpwalk on mib-2 tree takes 10 to 12 seconds. The same behaviour is observed on x86_64 architecture also.