Bug 231732
| Summary: | net-mgmt/net-snmp: hrSWRunEntry, don't list kernel threads | ||
|---|---|---|---|
| Product: | Ports & Packages | Reporter: | Ben RUBSON <ben.rubson> |
| Component: | Individual Port(s) | Assignee: | Ryan Steinmetz <zi> |
| Status: | Closed Overcome By Events | ||
| Severity: | Affects Some People | CC: | pi |
| Priority: | --- | Flags: | bugzilla:
maintainer-feedback?
(zi) |
| Version: | Latest | ||
| Hardware: | Any | ||
| OS: | Any | ||
Did you find a configure option or the like that accomplishes what you're after? No, I had to recompile the net-snmp package after having modified the sources with the command above.
It replaces the hard-coded op parameter to kvm_getprocs(), from KERN_PROC_ALL to KERN_PROC_PROC.
From KVM_GETPROCS(3) :
struct kinfo_proc * kvm_getprocs(kvm_t *kd, int op, int arg, int *cnt);
The value of op describes the filtering predicate as follows:
KERN_PROC_ALL all processes and kernel visible threads
KERN_PROC_PROC all processes, without threads
I think the best next steps for this are to submit a patch upstream to the net-snmp developers. Potentially making this a configure knob, if we think there are people that would want to see the kernel threads as well. Once they've committed something, I'm happy to backport it into the port. However, I'd prefer to avoid having additional locally-maintained-only patches. Thank you Ryan for your feedback, you're right. I then just opened an upstream bug / feature request : https://sourceforge.net/p/net-snmp/bugs/2890/ So, a patch has been pushed upstream, in 5.8 version. Current port version seems to be 5.7. https://www.freshports.org/net-mgmt/net-snmp/ We'll then be able to close this bug report once port updated to 5.8. Thank you ! The patch was added to upstream after the 5.8 release, so it's still pending. Handled as part of the 5.9 update |
Hi, When browsing hrSWRunEntry (1.3.6.1.2.1.25.4.2.1), many "kernel" processes are returned. I have servers with around 100 user processes, and 1000 kernel threads, making hrSWRunEntry SNMP requests very slow. KVM_GETPROCS(3) helped me to find the following solution : find /usr/ports/net-mgmt/net-snmp -type f \ -exec grep -q KERN_PROC_ALL {} \; -ls \ -exec sed -i "" 's/KERN_PROC_ALL/KERN_PROC_PROC/g' {} \; hrSWRunEntry requests now take around 0.5 second instead of sometimes more than 20 ! Could we then think about disabling kernel threads in hrSWRunEntry ? I see no reason to keep them. Thank you very much ! Ben