Hi folks,
I've got a problem with a large diskgroup creating a new volume:
When I run:
vxassist -g dactyldg -U fsgen make CGY_DATA56 28200m \
layout=stripe alloc=A00u2v10 A00u4v30 ,
I receive the error message:
vxvm:vxassist: ERROR: creating volume CGY_DATA56:
Configuration too large for disk group log
Here is the vxdg list output:
Group: dactyldg
dgid: 1086351079.2173.miranda
import-id: 0.2157
flags:
version: 90
detach-policy: global
copies: nconfig=4 nlog=4
config: seqno=0.94324 permlen=11877 free=7621 templen=2393 loglen=197
config disk fabric_2 copy 1 len=11877 disabled
config disk fabric_6 copy 1 len=11877 disabled
config disk fabric_9 copy 1 len=11877 disabled
config disk fabric_13 copy 1 len=11877 disabled
config disk fabric_18 copy 1 len=1303 disabled
config disk fabric_20 copy 1 len=5831 disabled
config disk fabric_22 copy 1 len=11877 disabled
config disk fabric_26 copy 1 len=11877 disabled
config disk fabric_30 copy 1 len=11877 disabled
config disk fabric_34 copy 1 len=11877 disabled
config disk fabric_39 copy 1 len=1303 disabled
config disk fabric_43 copy 1 len=1303 disabled
config disk fabric_47 copy 1 len=11877 state=clean online
config disk fabric_51 copy 1 len=11877 state=clean online
config disk fabric_55 copy 1 len=11877 disabled
config disk fabric_59 copy 1 len=11877 disabled
config disk fabric_63 copy 1 len=11877 state=clean online
config disk fabric_67 copy 1 len=11877 state=clean online
config disk fabric_72 copy 1 len=11877 disabled
config disk fabric_75 copy 1 len=11877 disabled
config disk fabric_80 copy 1 len=1303 disabled
config disk fabric_82 copy 1 len=5831 disabled
config disk fabric_87 copy 1 len=1303 disabled
config disk fabric_90 copy 1 len=1303 disabled
log disk fabric_2 copy 1 len=1799 disabled
log disk fabric_6 copy 1 len=1799 disabled
log disk fabric_9 copy 1 len=1799 disabled
log disk fabric_13 copy 1 len=1799 disabled
log disk fabric_18 copy 1 len=197 disabled
log disk fabric_20 copy 1 len=883 disabled
log disk fabric_22 copy 1 len=1799 disabled
log disk fabric_26 copy 1 len=1799 disabled
log disk fabric_30 copy 1 len=1799 disabled
log disk fabric_34 copy 1 len=1799 disabled
log disk fabric_39 copy 1 len=197 disabled
log disk fabric_43 copy 1 len=197 disabled
log disk fabric_47 copy 1 len=1799
log disk fabric_51 copy 1 len=1799
log disk fabric_55 copy 1 len=1799 disabled
log disk fabric_59 copy 1 len=1799 disabled
log disk fabric_63 copy 1 len=1799
log disk fabric_67 copy 1 len=1799
log disk fabric_72 copy 1 len=1799 disabled
log disk fabric_75 copy 1 len=1799 disabled
log disk fabric_80 copy 1 len=197 disabled
log disk fabric_82 copy 1 len=883 disabled
log disk fabric_87 copy 1 len=197 disabled
log disk fabric_90 copy 1 len=197 disabled
As you can see, all active log disks have a length of 1799 although
the loglen of the DG is at only 197 (from creation time). I have set down
the nlog value to 4, 2, and 1; the "loglen" value does not change (I hoped it
would like the "free" value does when smaller disks disappear from the list
of enabled config disks - but no... :( )
It's a little surprising to me that this phenomenon appears now, since I have
created some 55 volumes of the exactly same layout before, and I assume
strongly that the loglen value has not changed since then.
Does anyone know of a method to either increase the logsize or circumvent
the problem in some other way?
Any help appreciated,
thanks
Wolfgang Erasmi
