Bug 1813200
Summary: | gssproxy memory leak with rpc.gssd | ||
---|---|---|---|
Product: | Red Hat Enterprise Linux 8 | Reporter: | Yongcheng Yang <yoyang> |
Component: | gssproxy | Assignee: | Robbie Harwood <rharwood> |
Status: | CLOSED ERRATA | QA Contact: | anuja <amore> |
Severity: | high | Docs Contact: | |
Priority: | high | ||
Version: | 8.2 | CC: | dpal, fdvorak, fs-qe, ndehadra, plambri, rharwood, steved, swhiteho, xzhou |
Target Milestone: | rc | Flags: | pm-rhel:
mirror+
|
Target Release: | 8.3 | ||
Hardware: | Unspecified | ||
OS: | Linux | ||
URL: | https://github.com/gssapi/gssproxy/pull/6 | ||
Whiteboard: | |||
Fixed In Version: | gssproxy-0.8.0-19.el8 | Doc Type: | No Doc Update |
Doc Text: | Story Points: | --- | |
Clone Of: | 1809277 | Environment: | |
Last Closed: | 2021-05-18 14:42:07 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: | 1894575 | ||
Attachments: |
Comment 23
Robbie Harwood
2020-07-31 20:33:33 UTC
Created attachment 1703242 [details]
valgrind output with gssproxy-0.8.3-3.fc33
I have checked the reproducer on fedora33 (Fedora-Rawhide-20200802.n.0 from beaker). Please find the above attached output. (In reply to Yongcheng Yang from comment #24) > Created attachment 1703242 [details] > valgrind output with gssproxy-0.8.3-3.fc33 What was the valgrind command line did you use? (In reply to Steve Dickson from comment #27) > (In reply to Yongcheng Yang from comment #24) > > Created attachment 1703242 [details] > > valgrind output with gssproxy-0.8.3-3.fc33 > > What was the valgrind command line did you use? valgrind --log-file=valgrind.out.xxx --trace-children=yes --leak-check=full --track-origins=yes -v --show-reachable=yes rpc.gssd (In reply to Yongcheng Yang from comment #28) > (In reply to Steve Dickson from comment #27) > > (In reply to Yongcheng Yang from comment #24) > > > Created attachment 1703242 [details] > > > valgrind output with gssproxy-0.8.3-3.fc33 > > > > What was the valgrind command line did you use? > > valgrind --log-file=valgrind.out.xxx --trace-children=yes --leak-check=full > --track-origins=yes -v --show-reachable=yes rpc.gssd I ran the above valgrind command with rpc.gssd not using gssproxy (/etc/nfs.conf: use-gss-proxy=0) over night and got the following results: ==115828== LEAK SUMMARY: ==115828== definitely lost: 0 bytes in 0 blocks ==115828== indirectly lost: 0 bytes in 0 blocks ==115828== possibly lost: 0 bytes in 0 blocks ==115828== still reachable: 12,942 bytes in 44 blocks ==115828== suppressed: 88 bytes in 1 blocks ==115828== ==115828== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 0 from 0) So the interaction between rpc.gssd and gssproxy is causing the memory leaks. Yongcheng, would you mind verifying this case in your testing? (In reply to Steve Dickson from comment #30) ... > Yongcheng, would you mind verifying this case in your testing? Yes, have just checked that NO memleak found with "use-gss-proxy=0" set under "[gssd]" in /etc/nfs.conf. I.e. https://beaker.engineering.redhat.com/jobs/4467749 But why the nfs krb5 mounting can success even when rpc.gssd not using gssproxy. (I thought it's mis-configured before..) Btw, seemingly the usage of "use-gss-proxy" is not mentioned in the man page rpc.gssd(8). Created attachment 1710785 [details]
Entire valgrid log with gssproxy disabled
(In reply to Yongcheng Yang from comment #31) > (In reply to Steve Dickson from comment #30) > ... > > Yongcheng, would you mind verifying this case in your testing? > > Yes, have just checked that NO memleak found with "use-gss-proxy=0" set > under "[gssd]" in /etc/nfs.conf. > I.e. https://beaker.engineering.redhat.com/jobs/4467749 Thanks! > > But why the nfs krb5 mounting can success even when rpc.gssd not using > gssproxy. (I thought it's mis-configured before..) It is only required on the server since it replaced rpc.svcgssd(8) which was a bit buggy... It is optional on the client side. > > Btw, seemingly the usage of "use-gss-proxy" is not mentioned in the man page > rpc.gssd(8). This is probably a doc bug... Created attachment 1711485 [details] Entire valgrid log with gssproxy enabled using an Window AD Just for grins... I ran the set up in comment 39 using a Window's AD as the KDC... Doing a quick scan it appears the errors are similar which is not too surprising. Just trying to look at this from all angels Created attachment 1712420 [details]
Entire valgrid log with gssproxy enabled with gss_release_oid() added
Created attachment 1712422 [details]
Entire valgrid log with gssproxy disabled with gss_release_oid() added
Thanks, that first one is the log I was looking for. With that information, the correct fix is on the gssproxy side and you don't need to do anything for it: https://github.com/gssapi/gssproxy/pull/10 Verified with sanity run as per comment #68 and comment #67 Verfied bug using dev compose: gssproxy-0.8.0-19.el8.x86_64 ipa-server-4.8.7-13.module+el8.3.0+8376+0bba7131.x86_64 Adding report using bash::ipa-client-automount: 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] ====================================== Final Report ======================================= 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] Test Date: Wed Dec 2 03:52:40 EST 2020 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] Total : [57] 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] Passed: [57] 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] Failed: [0] 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] Unfinished: [0] 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] Abort : [0] 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] Crash : [0] 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] +-----------------------------[RPMs & OS: [RedHat - x86_64]-----------------------------+ 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] | ipa-client-4.8.7-13.module+el8.3.0+8376+0bba7131.x86_64 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] | ipa-client-common-4.8.7-13.module+el8.3.0+8376+0bba7131.noarch 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] | sssd-ipa-2.4.0-2.el8.x86_64 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] ------------------------------------------------------------------------------------------ 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] +-----------------------------------------------------------------------------------------+ 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] Test:[/ipa-server/rhel80/ipa-client-automount/root]: [ Pass(57/57): 100% ] 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] +-----------------------------------------------------------------------------------------+ 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] +----------------------------------------------------------------------+ 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] Fail / unfinished / ABORT [ Fail(0/57): 0% ] 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] +----------------------------------------------------------------------+ 2020-12-02T08:52:41+0000 [ci-vm-10-0-155-38.ho] =========================== end of report [/tmp/rhts.report.23479.txt]=============================== Based on this marking Verified:Tested SanityOnly Verified with sanity run as per comment #68 and comment #67 Adding report using bash::ipa-client-automount: 2020-12-17T07:17:46+0000 [ci-vm-10-0-107-23.ho] ====================================== Final Report ======================================= 2020-12-17T07:17:46+0000 [ci-vm-10-0-107-23.ho] Test Date: Thu Dec 17 07:17:45 UTC 2020 2020-12-17T07:17:46+0000 [ci-vm-10-0-107-23.ho] Total : [57] 2020-12-17T07:17:46+0000 [ci-vm-10-0-107-23.ho] Passed: [57] 2020-12-17T07:17:46+0000 [ci-vm-10-0-107-23.ho] Failed: [0] 2020-12-17T07:17:46+0000 [ci-vm-10-0-107-23.ho] Unfinished: [0] 2020-12-17T07:17:46+0000 [ci-vm-10-0-107-23.ho] Abort : [0] 2020-12-17T07:17:46+0000 [ci-vm-10-0-107-23.ho] Crash : [0] 2020-12-17T07:17:46+0000 [ci-vm-10-0-107-23.ho] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2020-12-17T07:17:46+0000 [ci-vm-10-0-107-23.ho] +-----------------------------[RPMs & OS: [RedHat - x86_64]-----------------------------+ 2020-12-17T07:17:46+0000 [ci-vm-10-0-107-23.ho] | ipa-client-4.9.0-0.5.rc3.module+el8.4.0+9124+ced20601.x86_64 2020-12-17T07:17:46+0000 [ci-vm-10-0-107-23.ho] | ipa-client-common-4.9.0-0.5.rc3.module+el8.4.0+9124+ced20601.noarch 2020-12-17T07:17:46+0000 [ci-vm-10-0-107-23.ho] | sssd-ipa-2.4.0-3.el8.x86_64 2020-12-17T07:17:46+0000 [ci-vm-10-0-107-23.ho] ------------------------------------------------------------------------------------------ 2020-12-17T07:17:46+0000 [ci-vm-10-0-107-23.ho] +-----------------------------------------------------------------------------------------+ 2020-12-17T07:17:46+0000 [ci-vm-10-0-107-23.ho] Test:[/ipa-server/rhel80/ipa-client-automount/root]: [ Pass(57/57): 100% ] 2020-12-17T07:17:46+0000 [ci-vm-10-0-107-23.ho] +-----------------------------------------------------------------------------------------+ Based on this marking verified. 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 (gssproxy bug fix and enhancement update), 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-2021:1592 |