Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations SkipVought on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Puredisk: incorrect available space and written data

Status
Not open for further replies.

allywilson

Technical User
Nov 27, 2002
157
0
0
GB
Hi all,
I'm new to Netbackup and Puredisk. Added our new puredisk pool to NB 6.5.3.

If you SSH to the puredisk server and 'df -h' it will show that we have available disk space 9.1TB. From the netbackup console, under Disk Pools it shows the Raw Size correctly, but it seems to think that 10% of the available size has been written to (nearly a terabyte - and I've only written 160MB to it!).

Is anyone aware of anything that might be causing this?
 
Hi, well I've no experience with pure disk so do not know how it presents itself as a file system to the unix server. BUT, if it follows the normal default rules then there is a 10% minimum overhead to UFS filesystems on Solaris servers.
This can be changed with tunefs -m <minfree>.

This may explain what you are seeing.

Hope this helps.

BTW, we are toying with the idea of De-dupe. How are you finding pure disk?

Thanks
 
We're looking to move to it in a big way, cost savings and all that.

Puredisk is a good product, unfortunately the lack of available documentation for it (especially its limitations!) has driven me insane.

It runs on top of SLES10 (which is cool), but the installation process is very unusual. Installing from their ISO doesn't actually install what you need, you need a second tar file which actually installs the puredisk application.

Oh, and forget about patching it or attempting to remove it cleanly. Messiest installation process I've come across in a long time. Tried both a number of times, and they bomb out - but don't seem to do a good dependancy check so you're left with half a system upgraded/removed!

I think I've eventually found out why my original issue is (will double-check tomorrow). It seems the VxFS filesystem is restricted in-size by its block size. 1024 = 4Tb, 2048 = 8Tb, 4096 = 16Tb.

So when I presented my 9TB storage pool to netbackup - VxFS could only handle up to 8TB as the blocksize is 2048.

I'll need to reformat the puredisk volume to use 4096 as the blocksize - fun and games...

Thanks for your response!

Ally
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top