Bug 860308
Summary: | condor SEGFAULT after upgrade while using custom hostname | ||
---|---|---|---|
Product: | Red Hat Enterprise MRG | Reporter: | Martin Bukatovic <mbukatov> |
Component: | condor | Assignee: | Timothy St. Clair <tstclair> |
Status: | CLOSED ERRATA | QA Contact: | Martin Bukatovic <mbukatov> |
Severity: | low | Docs Contact: | |
Priority: | low | ||
Version: | 2.2 | CC: | matt, mkudlej, sgraf, trusnak, tstclair |
Target Milestone: | 2.3 | ||
Target Release: | --- | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Fixed In Version: | condor-7.8.5-0.1 | Doc Type: | Bug Fix |
Doc Text: |
Cause:
Statically configuring /etc/hosts which may have ipv6 entries.
Consequence:
Schedd will crash trying to forward resolve entries based on the CVE fix.
Fix:
Properly handle ipv6 addresses in /etc/hosts and static configurations
Result:
Condor starts normally.
|
Story Points: | --- |
Clone Of: | Environment: | ||
Last Closed: | 2013-03-06 18:46:50 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: |
Description
Martin Bukatovic
2012-09-25 14:20:04 UTC
Interesting note: When I don't use ALL_DEBUG=D_FULLDEBUG option, I will get slightly different stack trace in the condor master log: ~~~ 09/25/12 16:30:55 WARNING: forward resolution of localhost4 doesn't match e77aa8c0! Stack dump for process 6657 at timestamp 1348583455 (19 frames) condor_master(dprintf_dump_stack+0x44)[0x810cfb4] condor_master[0x8144a87] [0x2c8400] /lib/libc.so.6(__nss_hostname_digits_dots+0x39)[0x549239] /lib/libc.so.6(gethostbyname+0x9a)[0x54e3ba] condor_master(_Z18verify_name_has_ipPc7in_addr+0x34)[0x80cc044] condor_master(_ZN8IpVerify6VerifyE12DCpermissionPK11sockaddr_inPKcP8MyStringS7_+0x6bd)[0x80ce69d] condor_master(_ZN6SecMan6VerifyE12DCpermissionPK11sockaddr_inPKcP8MyStringS7_+0x3d)[0x80e174d] condor_master(_ZN10DaemonCore6VerifyEPKc12DCpermissionPK11sockaddr_inS1_+0x71)[0x80a3871] condor_master(_ZN10DaemonCore9HandleReqEP6StreamS1_+0xcb7)[0x80b1c87] condor_master(_ZN10DaemonCore22HandleReqSocketHandlerEP6Stream+0x5f)[0x80b47bf] condor_master(_ZN10DaemonCore24CallSocketHandler_workerEibP6Stream+0x5af)[0x80b4f6f] condor_master(_ZN10DaemonCore35CallSocketHandler_worker_demarshallEPv+0x2d)[0x80b503d] condor_master(_ZN13CondorThreads8pool_addEPFvPvES0_PiPKc+0x57)[0x81429f7] condor_master(_ZN10DaemonCore17CallSocketHandlerERib+0x107)[0x80aae17] condor_master(_ZN10DaemonCore6DriverEv+0x1f6d)[0x80af99d] condor_master(main+0x1432)[0x809e4b2] /lib/libc.so.6(__libc_start_main+0xe6)[0x467ce6] condor_master[0x8092761] ~~~ Note that this one is the same as in 2 other bug reports I linked to above. When run without ALL_DEBUG=D_FULLDEBUG, it seems that condor crashes because of calling gethostbyname with wrong string - as can be seen in the following excerpt from ltrace log: ~~~ 6699 16:48:01.259591 write(10, "09/25/12 16:48:01 WARNING: forward resolution of localhost4 doesn't match e77aa8c0!\n ", 84) = 84 6699 16:48:01.259728 fflush(0x9e49510) = 0 6699 16:48:01.259827 fclose(0x9e49510) = 0 6699 16:48:01.259953 umask(022) = 022 6699 16:48:01.260061 sigprocmask(2, 0xbfb6d1dc, NULL) = 0 6699 16:48:01.260215 gethostbyname("\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\ 377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377 \377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\37 7\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\3 77\377\377\377\377\377\377\377\377\377\377"... <unfinished ...> 6699 16:48:01.260337 --- SIGSEGV (Segmentation fault) --- ~~~ Log was generated using: ltrace -tt -n 2 -f -s 120 -o condor_ltrace condor_master I was able to reproduce the problem on fresh virtual machine using the following steps: Steps to Reproduce: 1) install fresh rhel 6.3 2) change hostname to 'rhel-6-x86_64.virtualdomain' edit /etc/sysconfig/network 3) add following lines into /etc/hosts: # local pool 192.168.122.7 rhel-5-i386.virtualdomain rhel-5-i386 192.168.122.198 rhel-5-x86_64.virtualdomain rhel-5-x86_64 192.168.122.231 rhel-6-i386.virtualdomain rhel-6-i386 192.168.122.169 rhel-6-x86_64.virtualdomain rhel-6-x86_64 where 192.168.122.169 is global ipv4 address of the machine 4) reboot machine (for hostname to be updated) 5) install from mrg 2.1 install these packages: condor-7.6.5-0.14.el6.x86_64.rpm condor-classads-7.6.5-0.14.el6.x86_64.rpm condor-debuginfo-7.6.5-0.14.el6.x86_64.rpm 6) start condor if it's not already running, run condor_status to see that it's working, stop condor 7) upgrade to mrg 2.2 (just do yum upgrade) 8) start condor, try condor_status, see logs 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. http://rhn.redhat.com/errata/RHSA-2013-0564.html |