[K12OSN] Uninteruptable Power Supply UPS with Linux

Tue Oct 17 03:07:38 UTC 2006

On Mon, 2006-10-16 at 13:04 -0600, Calvin Dodge wrote:
> On 10/16/06, Robert Arkiletian <robark at gmail.com> wrote:
> >
> >
> >  Why do you recommend usb over serial?
> 
> The cheaper APCs don't do true serial communication (i.e., sending
> data to the computer). Instead, they set certain signals (like CTS,
> IIRC) to indicate their condition.

It even more complicated than that. The different models transmit
differing amounts of data. Whether the data port is serial or usb makes
no difference. The Back-UPS vs the Smart-UPS have a large amount of
functionality difference. The Backups is simpler and has only a timer
function. Once it goes on battery, the timer starts. At the end of the
timer, it send the shutdown command to the host.

The smartups has much more of a command set. It supports battery
condition testing and and can signal a power down when the battery falls
below a certain level. More importantly, it can hold of on powering back
up until the battery has reached a certain minimum runtime level.

On my current install, we have Smart-UPS XL 3000VA RM 3U models that
have both serial and USB connectors. The same data is on both ports.

One nice thing about these (and all Linux systems can be setup to do
this) is when the shutdown command is issued, the apcupsd can be told to
also send it to a group of other machines. In my current case, there are
several servers that run off the UPS in the rack. If the UPS is going
down that powers the main login/NFS-/home machine, every other server
will need to also shut down as they will have nothing they can do. If a
UPS signals to a boot/app server the remaining app servers can stay up.

So the UPS is told power back up when the AC as back and the batteries
are back up to enough time to allow a partially booted system to shut
down gracefully. The first box up is the NFS server. It will then tickle
the other machines with Wake-on LAN calls (as soon as I can finish that
process).
> 
> USB allows for real communication with the host computer, and I
> _believe_ that means real data gets sent to that host. For example,
> here's the output of "apcaccess" on my workstation:
> 
> APC      : 001,034,0890
> DATE     : Mon Oct 16 12:59:50 MDT 2006
> HOSTNAME : source-server.prosocial.local
> RELEASE  : 3.12.4
> VERSION  : 3.12.4 (19 August 2006) redhat
> UPSNAME  : source-server.prosocial.local
> CABLE    : USB Cable
> MODEL    : Back-UPS NS 600
> UPSMODE  : Stand Alone
> STARTTIME: Fri Oct 13 13:05:46 MDT 2006
> STATUS   : ONLINE
> LINEV    : 118.0 Volts
> LOADPCT  :   0.0 Percent Load Capacity
> BCHARGE  : 100.0 Percent
> TIMELEFT :  42.8 Minutes
> MBATTCHG : 5 Percent
> MINTIMEL : 3 Minutes
> MAXTIME  : 0 Seconds
> LOTRANS  : 088.0 Volts
> HITRANS  : 139.0 Volts
> ALARMDEL : Always
> BATTV    : 13.5 Volts
> LASTXFER : Low line voltage
> NUMXFERS : 0
> TONBATT  : 0 seconds
> CUMONBATT: 0 seconds
> XOFFBATT : N/A
> STATFLAG : 0x07000008 Status Flag
> MANDATE  : 2005-12-26
> SERIALNO : 3B0601X00177
> BATTDATE : 2000-00-00
> NOMBATTV :  12.0
> FIRMWARE : 18.w1 .D USB FW:w1
> APCMODEL : Back-UPS NS 600
> END APC  : Mon Oct 16 12:59:51 MDT 2006
> 
> 
> I'm reasonably certain a serial signal-only connection can't convey
> that much information to the host (I've no great emotional stake in
> this, and won't complain if somebody can post similar output from a
> signal-only UPS).
> 
> >
> >  I am using the CentOS 4 based k12ltsp.
> 
> RPMs are available for RHEL3 at the apcupsd site
> (http://sourceforge.net/project/showfiles.php?group_id=54413&package_id=73150).
> I think that's the closest match to CentOS 4 (being directly descended
> from RHEL4).
> 
> Calvin
> 
> _______________________________________________
> K12OSN mailing list
> K12OSN at redhat.com
> https://www.redhat.com/mailman/listinfo/k12osn
> For more info see <http://www.k12os.org>
-- 
James P. Kinney III          
CEO & Director of Engineering 
Local Net Solutions,LLC        
770-493-8244                    
http://www.localnetsolutions.com

GPG ID: 829C6CA7 James P. Kinney III (M.S. Physics)
<jkinney at localnetsolutions.com>
Fingerprint = 3C9E 6366 54FC A3FE BA4D 0659 6190 ADC3 829C 6CA7
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://listman.redhat.com/archives/k12osn/attachments/20061017/2b1fd9c6/attachment.sig>