SNMP & SNAPs

[rusty.hill]

Hi

I have a bunch of Snaps which I've upgraded over the years - the bigest is 1.5TB via a pair of 750GB Seagates in a 2000 - and use SNMP to manage the beasts.

My question is, does anyone have a SNAP MIB which I can use to import into my NetMon tool?

The SNAP 1000/1100/2000/2200 series support some of the basic RFC SNMP functions such as uptime, interface performance etc but I'm intreged to see that if I catch SNMP traps I get more info- such as overheated cpu, fax failure etc.

Now traps are fine but a MIB which I can interogate is better! I can't find either the SNAP/Quantum/Adaptec OID list or MIB -can anyone help pls?

BTW, the current Adaptec 'general storage' MIB does not 'know' about Snaps.

TiA

Rusty Hill

[wgong]

Originally Posted by rusty.hill

Hi
My question is, does anyone have a SNAP MIB which I can use to import into my NetMon tool?

My 705N (4100) doco does not list a MIB, only differences between the standard MIBs and what is implemented by the NAS.
Mentioned are: RFC1213 (MIB-II), RFC1514 (Host Resources) and Proprietry (DellPV705N).

The SNAP 1000/1100/2000/2200 series support some of the basic RFC SNMP functions such as uptime, interface performance etc but I'm intreged to see that if I catch SNMP traps I get more info- such as overheated cpu, fax failure etc.

Here is what the 705N NAS Technical Reference says about traps:

If enabled, SNMP traps occur whenever there is an event on the PowerVault NAS Appliance that is at an 'error' level or above. The implementation of this is similar to the current email notification logic. Essentially, if traps are enabled, a trap is generated whenever an error occurs.

The following traps are possible:

* Cold start trap - it is sent out at boot time.
* Warning-level traps - several warning-level traps are sent out when the specific warning condition is encountered, for example, when available disk space is too low.
* Error-level traps - an error-level trap is sent out for each PowerVault NAS Appliance error.
* Fatal Error-level traps - an attempt is made to send a fatal error trap for each fatal error. In some instances this is not possible.
* Product agent status change trap - it is sent when the global agent status changes, for example from OK to Error.
* Executing Fsck - it is sent when Fsck is executed on the server.
* Partition is clean - it is sent after Fsck is executed and the partition is found to be clean.

I will have a look through Dell's OpenManage MIBs and see if the 701N/705N specific MIBs are listed.

[rusty.hill]

Thanks for that - interesting.

I see the use of the 'Enterprise' MIB for the Dell701... I guess that might help with the 1000/2000 series as ISTR that Dell 701 is really made/was made by SNAP & just had a sticky Dell badge on it.

I'll see if I can find a copy of the mib on the 'net.....

Thanks again,

Rusty

[dgoodrich]

Rusty.hill,

Sorry, I can't help you on the MIB (since I don't know what that is)

I am very interested in your 1.5TB 2000, however. Did you do anything special with the power supply or the fan(s) to accomodate the large drives?

Thanks,

[rusty.hill]

Hi

Doh myself for not explaining it...

MiB = Managment Information Base - see:

http://en.wikipedia.org/wiki/Managem...formation_base


Using SNMP - simple network monitoring protocol one can get devices to both send alerts to a central monitoring PC as well as letting the monitoring PC poll the devices to see what they are doing.

By RFC all 'grown up' devices have unique MiBs and the makers usually give these data files away with the devices or on their web sites. If you don't have the right MiB for the switch, server, router etc all you can get is a string of numbers back rom it - with the right MiB you can automagically convert the string into an English error message - viz: 'I'm too hot!! - my fan has failed!' etc.

Which, to cut a long story short, is how I knew the cpu fan on one of my 2000's had failed as I got a 'I'm too hot' message.....

and yes, on the 2000/2200's I've replaced the cpu fans with a bigger fan to keep them cool. (Remember this fan only runs when required so may take 5 mins or so to kick in after power up.)

So far I've not replaced the main psu cooling fans but it would be worth doing as cooler is always better!

See my notes in the Procooling Wiki for details on putting 1.5TB and 1TB in Snap 2000's - good luck!

Brgds

Rusty

[dgoodrich]

Thanks for the info. I hadn't noticed your section in the Wiki before. Its cool to see a 2000 can accomodate drives that big with only minor modifications. I've noticed that with (2) 250-gig WD 7200 rpm drives, the air coming out of the chassis during heavy transfers barely seems warm, so I haven't worried about it much.

So the factory power supply is able to spin up those big drives without a problem?

Thanks again,
Dave