Bug 1196320
| Summary: | lldpad memory usage increases over time | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Product: | Red Hat Enterprise Linux 7 | Reporter: | Stuart Auchterlonie <sauchter> | ||||||
| Component: | lldpad | Assignee: | Aaron Conole <aconole> | ||||||
| Status: | CLOSED ERRATA | QA Contact: | Ondrej Mejzlik <omejzlik> | ||||||
| Severity: | urgent | Docs Contact: | |||||||
| Priority: | urgent | ||||||||
| Version: | 7.0 | CC: | aconole, cleech, coughlan, dbragalo, jeharris, jmagrini, kcleveng, msugaya, omejzlik, ptalbert, qe-baseos-daemons, redhat, revers, ryamamot, sukulkar, toneata, yanwang | ||||||
| Target Milestone: | rc | Keywords: | OtherQA, Triaged, ZStream | ||||||
| Target Release: | --- | ||||||||
| Hardware: | Unspecified | ||||||||
| OS: | Unspecified | ||||||||
| Whiteboard: | |||||||||
| Fixed In Version: | lldpad-1.0.1-5.git036e314 | Doc Type: | If docs needed, set a value | ||||||
| Doc Text: | Story Points: | --- | |||||||
| Clone Of: | |||||||||
| : | 1275488 1683960 1683961 (view as bug list) | Environment: | |||||||
| Last Closed: | 2019-08-06 13:18:45 UTC | Type: | Bug | ||||||
| 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: | 1203710, 1400961, 1420851, 1472751, 1477664, 1683960, 1683961 | ||||||||
| Attachments: |
|
||||||||
One further bit of information from the customer, every server that they saw this happening on, is actually a VM running on an ESXi hypervisor. So they shouldn't really be running lldpad on these servers, but equally lldpad should not leak memory just because it's running on a VM. I am seeing the same problemm also on a ESXi hypervisors. It appears to be trying to access something that doesn't exist on VMs and loops about once per second trying to access the resource. Every time it loops, it adds about 5 bytes to the memory footprint. These are El6 and El7 hosts. Created attachment 1233603 [details]
Strace output of a lldpad process that is malfunctioning.
Several of the subtype modules look to have incorrect return values when encountering unexpected TLVs. It makes it look as if they have taken ownership of an allocation, and if nobody else actually claims it then the memory will be leaked. I'm working up a test build, will share it when it's ready. Test build for x86_64 RPMS at http://people.redhat.com/cleech/bz1196320/ Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2019:2339 *** Bug 1513811 has been marked as a duplicate of this bug. *** |
Created attachment 995271 [details] Output of /proc/<pid>/smaps for lldpad consuming excess memory Description of problem: lldpad memory usage increases over time Version-Release number of selected component (if applicable): lldpad-0.9.46-2.el6.x86_64 lldpad-libs-0.9.46-2.el6.x86_64 How reproducible: 100% by customer Steps to Reproduce: 1. run lldpad for an extended period of time 2. 3. Actual results: Memory usage of lldpad increases Expected results: Memory usage of lldpad should remain stable over time Additional info: See lines 44-46 of smaps data