multipath path flapping for hours after a one minute downtime
Issue
- During firmware upgrades of SAN switches that when we lose a path, the path flaps between active and failed for many hours (up to 10+) afterwards.
- Also, during this time, performance on the host is horrible. Commands hang, backups fail, etc..
Environment
- RHEL 6.5 kernel version 2.6.32-431.29.2.el6.x86_64
- device-mapper-multipath 0.4.9-72.el6_5.4.x86_64
- EMC SYMMETRIX (Rev: 5874) external storage
- Cisco UCS 6296UP switch running 5.0(3)N2(2.11.3c)
- CISCO UCS blades with VIC FCoE HBAs (rev a2)
- Default/built-in multipath.conf settings for SYMMETRIX (using TUR path_checker)
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.