Hi All,

I am writing a pseudo driver for which a want to add support for sun
cluster. I am really puzzle to see the way devices are getting
accessed on the secondary node (node where resource group is offline).

I have setup a two node fail-over sun cluster and did some experiments
to understand the device status on different cluster nodes. one of
them which seems bizarre to me are as follows:-

1. I have mounted file-system on clustered device on the primary
without global option.
   Now what is my understanding is on other node this device should
not be accessible in the read/write mode (or at least in write
   mode) since the device is mounted on the primary. But I could
mount (without global option) the same device even on secondary
   node and not only that I create some files as well. Even I could
wrote to the device using dd(1m) no wonder it corrupted the
   file-system.

Now, my question is why shared devices is not protected from secondary
while is mounted on primary node?
if this is expected behavior that means sun cluster entirely relies on
application of behave well and if someone by mistake do a dd(1m) on
the device from the secondary then everything is gone for the good, is
that acceptable behavior?

May be there could be configuration issue in my cluster that why I was
allowed to access the device from the secondary. I am using HA storage
plus and default fencing is pathcount and the file-system type was
ufs.

Any idea or suggestion would be a great rescue.

thanks,
ajit jain

PS: please add in cc list, I am a member of the mailing list.

Reply via email to