Hello,
I am using two log plexes in my vxvm 3.5 RAID 1 mirrored volume. Both log
plexes are on the same LUN as the concat mirror. i.e
# vxprint -ht -g oraldg2 -v admin | grep L6C
sd BE2C00L6C-02 admin-01 BE2C00L6C 41944320 16778880 0 HDS9980V1_4
ENA
sd JM2C00L6C-03 admin-02 JM2C00L6C 41946240 16778880 0 HDS9980V0_5
ENA
sd JM2C00L6C-04 admin-03 JM2C00L6C 363 165 LOG HDS9980V0_5
ENA
sd BE2C00L6C-06 admin-04 BE2C00L6C 92279700 165 LOG HDS9980V1_4
ENA
#


When my cluster (sc3.1u1) switches, I sometimes get the following error message:
-
"Cluster.Framework: [ID 801593 daemon.error] stderr: vxvm:vxvol: ERROR:
Volume admin: Invalid recovery log magic number"
The concat mirrors themselves remain intact and the ufs filesystem mounts
successfully. I simply have to reset (i.e remove and re-add) one of the
DRLs.

Is there any reason (excluding performance) why I should not have two DRLs
and locate them on the same disk as the concat plexes.

Rgds,
Colin