[OpenIndiana-discuss] Disk Space Disappearing.

[Dormition Skete (Hotmail)]

On Jul 30, 2014, at 8:20 PM, Timothy Coalson <tsc5yc at mst.edu> wrote:

> USED is the sum of all the following USED* properties - USEDSNAP is the
> space that would be freed by deleting all the snapshots on this filesystem,
> USEDDS is amount of live data on this filesystem, USEDCHILD is amount used
> (or reserved) by child filesystems, and USEDREFRESERV is amount of reserved
> space for this filesystem that isn't actually used by anything (I think…).

Thank you.  That makes more sense now.



>> Is there any way to tell why the swap is so high?  This machine has loads
>> of RAM.  Or is that normal for Solaris?
>> 
> 
> As I understand it, the swap zvol is always preallocated, and default size
> appears to be roughly half the available RAM at install time.  This does
> not mean any swap is in use.
> 

That’s good to hear.  I have 16 GB of RAM in the machine.  (I think.)



>> And does anything else stand out to any of you?
>> 
>> I see these:
> 
> NAME                                     AVAIL   USED  USEDSNAP  USEDDS
> USEDREFRESERV  USEDCHILD
> ...
> rpool/zones/archive/ROOT/zbe-2           37.1G  52.4G     24.3G   28.1G
>         0          0
> ...
> rpool/zones/routerb2/ROOT/zbe-2          37.1G  16.3G     7.16G   9.16G
>         0          0
> 

These are our two email zones.  Our Dovecot server is in routerb2 because I couldn’t get mail to work unless it had a direct public IP address.  So the zone’s name does not really reflect that.  All of our other zones go through that routerb2 zone to get their internet access.  The “archive” zone is our email archives for the past 17 years.  

We have a LOT of email.  Especially when you consider there’s only four of us here.


>> 
>> 
>>> I see "10M", not "10G", so I'm going to say no, it won't free that much
>>> space.
>> 
>> 
> 
> I hadn't looked at that, but I think most of it is unrelated to the boot
> environment (only the first line is obviously a snapshot on the boot
> environment).  A shot in the dark here- is "2014-07-05-23:13:59" the format
> of the snapshot names your backup script uses?  If so, and that date/time
> matches up with a backup you did, you probably made those snapshots
> yourself.
> 

I read someplace — I think on Oracle’s website — that when you create a boot environment, it creates a snapshot of the various zones, labeling them zbe-1, zbe-2, etc. for each new BE.  

I have deleted BE’s in the past, though not on a production server, without any problems.  

Having had four out of five servers, and a telephone system, all have hardware failures of one sort or another this past month, and every time I turned around, I was just breaking things worse... let’s just say I’m a little worn out by all of this.  (Actually, all five servers have given me grief.)  All I’ve done this month is wrestle with these machines...

So again, thank you very, very much for all of your help.  I really appreciate it a lot.