Systems configured with a large number of LVM PVs may boot to emergency mode.

This document (7021232) is provided subject to the disclaimer at the end of this document.

Environment

SUSE Linux Enterprise Server 12 SP4

SUSE Linux Enterprise Server 12 SP3

SUSE Linux Enterprise Server 12 SP2

SUSE Linux Enterprise Server 12 SP1

Situation

The system boots to emergency mode after a large delay.

Resolution

The LVM community is currently aware of this issue. Until the time required for PVs to be scanned can be improved it is recommended that lvmetad be disabled in such configurations.

This may be accomplished by changing the use_lvmetad value in /etc/lvm/lvm.conf from 1 to 0 and restarting the system. This is currently the default for SLES 12 SP1. For SLES 12 SP2 however, the default is currently set to 1 and should be change to 0 if this issue is encountered.

Currently the use of very high numbers of PVs is not recommended when using lvmetad.

Cause

With lvmetad active, the time required to activate volume groups (VGs) increases with each additional physical volume (PV) added to the system. A large enough number of PVs may cause the time required for a VG to be activated to exceed the timeout value of mount units associated with logical volumes (LVs) contained within that VG.

To check for this condition you may use systemd-analyze with the blame option to look for lvm2-pvscan units running for large periods of time:

# systemd-analyze --no-pager blame

2min 57.166s lvm2-pvscan@94:177.service

2min 53.540s lvm2-pvscan@94:245.service

Disclaimer

This Support Knowledgebase provides a valuable tool for SUSE customers and parties interested in our products and solutions to acquire information, ideas and learn from one another. Materials are provided for informational, personal or non-commercial use within your organization and are presented "AS IS" WITHOUT WARRANTY OF ANY KIND.