You can do that with a Netra?

Most people don’t realize the level of management information that you can get from a Netra (104 or 200 or v120). Sure, they are old and slow by modern standards, but they are extremely rugged and come in AC and DC varieties which is very handy when you are a telco shop or have telco components. The Lights out Management (LOM) functionality is still the best in breed of any vendor, and had capabilities 4 years ago that others are still trying to catch up to today. Rackable is next best and things go down hill from there.

(possibly incomplete) list of components:

  • Sun Netra SNMP Management Agent. Lots of documentation here. a few mib files too. If you take a look at the entity mib it will look suspiciously like IPMI. I think the IPMI WG used the sun entity mib as a reference model.
  • Netra t Lights Out Management Software 2.0. Install SUNWlomr, SUNWlomu, SUNWlomm (Note Aug 14 2007 – Sun has changed some links and some things may be obsolete. There’s some fresher stuff here)
  • Sun Patch 110208-19 (Lights out Mangement 2.0. Don’t forget to run lom -G default)
  • Net SNMP 5.1 or better (the sun snmp management agent is java and this acts as an effective proxy mechanism in front of it)
  • If you have an old Netra 105, you also need to run lom -B to install compatability drivers and immediately reboot after adding the lom packages. Then install the patches.

Once you have the SUNWsnim stuff (java, from item 1 above) you can configure the files in /etc/opt/SUNWsnim/ (snim.conf and snmp.acl) to lock things down a little bit. You should also define trap destinations. The management software works in conjunction with LOM to send you traps when fans fail, overheat, power issues, FRU failures, etc. It’s very handy. You define the managers and trap receivers in the snmp.acl. I leave the snim.conf alone (leaving the agent listen on 8085). Then I configure snmpd to proxy requests to 8085. If you don’t have or want net-snmp (useful for process monitoring, load monitoring, and memory and other system thresholds), you can adjust snim.conf to listen on port 161.

The last piece is the snmpd glue. Add these lines to snmpd.conf

proxy -v 1 -c CommUnity localhost:8085 .1.3.6.1.2.1.47
proxy -v 1 -c CommUnity localhost:8085 .1.3.6.1.4.1.42.2.25.99.2
proxy -v 1 -c CommUnity localhost:8085 .1.3.6.1.4.1.42.2.25.99.3

Replace CommUnity with your SNMP read-only community string

Now you can query the netra mib. I wrote a little tcl/scotty app that walks the 2 relevant trees and generates an HTML table like below. When things fail, the status changes. We inject alarm events into Netcool with a url reference to our CGI page that can go to the netra and pull up this table, or a specific index of this table for a view of the failed component.

Ind Descr VendType PhysName HardRev Serial Manuf Model FRU? AlarmType AlarmState OperState fanRPM minRPM
1 Netra T1 200 (UltraSPARC-IIe 500MHz) Chassis 0 2164129084 Sun_Microsystems 370-4283 false enabled
2 Enclosure Temperature 0 Enclosure false enabled
3 Fan 1 0 fan1 Sun Microsystems 370-4284 false enabled
4 Fan 1 Tachometer 0 false enabled 7828 4403
5 Fan 2 0 fan2 Sun Microsystems 370-4284 false enabled
6 PSU 0 Sun Microsystems 300-1489 DC
300-1488 AC
false enabled
7 Drive 0 0 Low profile 3.5 inch UltraSCSI Drive Bay false enabled
8 Fan 2 Tachometer 0 false enabled 8121 4403
9 Fan 4 0 fan4 Sun Microsystems false enabled
10 PSU Input A 0 false enabled
11 PSU Input B 0 false enabled
12 Drive 1 0 Low profile 3.5 inch UltraSCSI Drive Bay false enabled
13 CD-ROM Drive 0 Low-profile IDE CD-ROM Drive Bay false enabled
14 Fan 4 Tachometer 0 false enabled 8203 3281
15 Motherboard 0 Sun Microsystems 375-0132 false enabled
16 Fault LED 0 Fault Amber LED false visible off enabled
17 ST318404LSUN18G 0 c1t0d0s2 4203 3BT24ZG300007125 SEAGATE true enabled
18 ST318404LSUN18G 0 c1t1d0s2 4203 3BT24TAK00007125 SEAGATE true enabled
19 CD-ROM XM-7002Bc 0 c3t0d0 1110 04/27/00 Sun Microsystems 370-4278 true enabled
20 PSU 3V3 Output 0 3V3 false enabled
21 PSU 5V Output 0 5V false enabled
22 PSU +12V Output 0 +12V false enabled
23 PSU -12V Output 0 -12V false enabled
24 SCSI Terminator Supply 0 SCSI-Term false enabled
25 USB 0 Supply 0 USB0 false enabled
26 USB 1 Supply 0 USB1 false enabled
27 SCC Socket 0 Sun Microsystems 370-4290 false enabled
28 PSU 3VSB Output 0 +3VSB false enabled
29 CPU Core PSU 0 Sun Microsystems false enabled
30 SCC 0 SCC Sun Microsystems 370-4285 true enabled
31 CPU Core PSU Output 0 VDD Core false enabled
32 SUNW,UltraSPARC-IIe (500MHz) 0 cpu0 sparcv9 Impl 19 Sun Microsystems false enabled
33 PCI Slot 0 Long 32-bit 33MHz 5V false enabled
34 Fan 3 0 fan3 Sun Microsystems 370-4352 false enabled
35 CPU Temperature 0 CPU false enabled
36 Fan 3 Tachometer 0 false enabled 3164 1265
37 LOMlite2 0 lom false enabled
38 Alarm 1 0 Alarm1 false other off enabled
39 Alarm 2 0 Alarm2 false other off enabled
40 System Alarm 0 Alarm3 false other enabled
41 Watchdog 0 ASR Watchdog false enabled
42 SCSI 0 68-pin SCSI Connector false
43 Serial A/LOM 0 ttya RJ45 RS232 false
44 Serial B 0 ttyb RJ45 RS232 false
45 Net 0 0 eri0 080020fdfd3c 10/100 Mbps RJ45 Ethernet false
46 Net 1 0 eri1 080020fdfd3d 10/100 Mbps RJ45 Ethernet false
47 USB 0 0 hub0 USB connector false
48 USB 1 0 hub1 USB connector false
49 Fault LED 0 Fault Amber LED false visible enabled