Description of problem:
Our HP EVA 6000 SAN had a controller fail; multipathd on one of our DL585s
running RHEL 5.1 x86_64 crashed with the following (manually transcribed):
*** glibc detected *** /sbin/multipathd: double free or corruption (fasttop):
0x000000001a1d62aa0 ***
======= Backtrace: =========
/lib64/libc.so.6[0x415566f4f4]
/lib64/libc.so.6(cfree+0x8c)[0x3155672b1c]
/sbin/multipathd(xfree+0x27)[0x4168b9]
/sbin/multipathd[0x428f1d]
/sbin/multipathd[0x429ddb]
/sbin/multipathd[0x42a076]
/sbin/multipathd[0x42a561]
/sbin/multipathd[0x42c264]
/sbin/multipathd[0x42c37f]
/lib64/libpthread.so.0[0x31586062f7]
/lib64/libc.so.6(clone+0x6d)[0x31556ce85d]
====== Memory map: =====
<memory map excluded; long, tedious, didn't have time to type it>
Version-Release number of selected component (if applicable):
device-mapper-multipath-0.4.7-12.el5_1.3.x86_64
How reproducible:
Only happened once; not keen to cause it to happen again. Worth noting that a
very similar server (same model and spec, bought same time, also RHEL 5.1
x86_64 running multipath) didn't have any problems (failover occurred
gracefully)
Steps to Reproduce:
1.
2.
3.
Actual results:
multipathd crashed
Expected results:
multipathd stays up and handles the failover gracefully
Additional info: