Why SANscreen shows what it shows for switch health status?

You may be looking at SANscreen, and wondering why SANscreen shows what it shows for switch health status.

Here are the 3 main scenarios people ask about why it do what it do:

#1. You have Cisco, and we show nothing. This is a gap on our support matrix - currently, we just don't retrieve this for Cisco switches. Maybe if the Acquisition team gets an intern this summer, we can sic them on this.

#2. You have a non-Cisco, and we show something. You go to the device, and the device shows the opposite.

In SANscreen 0 - 5.1.x, switch health status is a minor change. If this is the *only* thing that has changed on the switch, SANscreen will not report the change - this was an internal decision we made to reduce the number of writes to the database, and to protect against "flickering" (imagine you have a wonky power supply that flip flops between good and bad. SANscreen polls this switch every 20 minutes - we would report 72 change events for the switch (3 per 24 hours) every day until this wonky power supply was replaced).

These minor changes *will* be reported when a major change occurs on the switch (SANscreen defines as something that would change paths - port up/down, zoning changes, etc).

The "good" news is that this will not be the case in 6.0 and higher - details we treated as minor changes in the past will be reported immediately.

#3. You have a Brocade, and you want to know why we show what we show.

We show... what Brocade tells us, converted to SANscreen-ese. Brocade FOS seems to have very very sensitive parameters for reporting a switch as "Marginal"

Here is an example for a Brocade switch discovered by a Brocade CLI datasource on a 5.1.4 SANscreen lab box here in scenic Waltham, MA:

- <RESPONSE exitCode="0">
  <STANDART_OUT>Switch Health Report Report time: 06/03/2010 01:54:57 PM Switch Name: ABCDEFGHIJK IP address: 10.60.345.67 SwitchState: MARGINAL Duration: 714:48 Power supplies monitor MARGINAL Temperatures monitor HEALTHY Fans monitor HEALTHY Flash monitor HEALTHY Marginal ports monitor HEALTHY Faulty ports monitor HEALTHY Missing SFPs monitor HEALTHY Fabric Watch is not licensed Detailed port information is not included</STANDART_OUT>
  <ERROR />

In the recording file for Brocade CLI,you can see the CLI commands we issue, and their response. We simply take the overall switch health "SwitchState: MARGINAL" , map it to our heterogenous model:

In the odata for this switch:


"4" is a code for what will appear in SANscreen as "Warning". "5" is "Faulty". "3" is "OK" . Basically, SANscreen has a set list of what we will display, and we map various vendor  terms to our list - Brocade's "Marginal" may be "Under the Weather" on a McData switch.

Hopefully this makes some sense. I should do a post on exploring SANscreen recordings (the zip files in ../sanscreen/acq/log folder)


Re: Why SANscreen shows what it shows for switch health status?

SANscreen 6.0.1 will have Cisco switch health status support. Woot

Re: Why SANscreen shows what it shows for switch health status?

Great info. thanks for taking the time to do this.