1062396 - remove tegra infrastructure from nagios

Status: RESOLVED FIXED

(Infrastructure & Operations :: RelOps: General, task)

Description

[Amy Rich [:arr] [:arich]]

In preparation for their decommissioning next week (and to avoid doing unnecessary work for bug 1053723) the releng tegra infrastructure (foopies and boards) were removed from nagios in commits:

r92823
r92824
r92825

We will also need to remove the switches from nagios, but want dcops' help identifying the correct ones to remove.

Comment 1

[Amy Rich [:arr] [:arich]]

dcops: from what I can see, it looks like these switches and PDUs are tegra only and should be removed from nagios (and eventually either decommed or marked as spare once we officially decomm the tegras):

Rack: 601-11:
switch1.r601-11.console.scl3.mozilla.net
switch2.r601-11.console.scl3.mozilla.net
pdu1.r601-11.tegra.releng.scl3.mozilla.com
pdu2.r601-11.tegra.releng.scl3.mozilla.com
pdu3.r601-11.tegra.releng.scl3.mozilla.com

Rack: 602-11:
switch1.r602-11.console.scl3.mozilla.net
switch2.r602-11.console.scl3.mozilla.net
pdu1.r602-11.tegra.releng.scl3.mozilla.com
pdu2.r602-11.tegra.releng.scl3.mozilla.com
pdu3.r602-11.tegra.releng.scl3.mozilla.com

Rack: 201-11:
switch1.r201-11.console.scl3.mozilla.net
switch2.r201-11.console.scl3.mozilla.net
pdu1.r201-11.tegra.releng.scl3.mozilla.com
pdu2.r201-11.tegra.releng.scl3.mozilla.com
pdu3.r201-11.tegra.releng.scl3.mozilla.com

Rack: 202-11:
switch1.r202-11.console.scl3.mozilla.net
switch2.r202-11.console.scl3.mozilla.net
pdu1.r202-11.tegra.releng.scl3.mozilla.com
pdu2.r202-11.tegra.releng.scl3.mozilla.com
pdu3.r202-11.tegra.releng.scl3.mozilla.com


The other two racks that contain foopies seem to also have other gear in them, so nagios monitoring should stay for those:
Rack: 202-9
Rack: 202-10

Did I miss anything (one of you can answer needinfo for all three)?

Comment 2

[Van Le [:van]]

That is correct. You didn't miss anything as far as I can tell.

Comment 3

[Amy Rich [:arr] [:arich]]

For some reason we have both console.scl3.mozilla.net addresses and ops.releng.scl3.mozilla.net addresses in releng/scl3.pp:

switch1.r602-11.console.scl3.mozilla.net
switch1.r201-11.console.scl3.mozilla.net
switch1.r202-11.console.scl3.mozilla.net
switch1.r601-11.console.scl3.mozilla.net
switch2.r201-11.console.scl3.mozilla.net
switch2.r202-11.console.scl3.mozilla.net
switch2.r601-11.console.scl3.mozilla.net
switch2.r602-11.console.scl3.mozilla.net
switch1.r602-11.ops.releng.scl3.mozilla.net
switch2.r602-11.ops.releng.scl3.mozilla.net
switch1.r601-11.ops.releng.scl3.mozilla.net
switch2.r601-11.ops.releng.scl3.mozilla.net
switch1.r202-11.ops.releng.scl3.mozilla.net
switch2.r202-11.ops.releng.scl3.mozilla.net
switch1.r201-11.ops.releng.scl3.mozilla.net
switch2.r201-11.ops.releng.scl3.mozilla.net

They also make an appearance in hosts/scl3.pp:
switch1.r602-11.console.scl3.mozilla.net
switch1.r201-11.console.scl3.mozilla.net
switch1.r202-11.console.scl3.mozilla.net
switch1.r601-11.console.scl3.mozilla.net
switch2.r201-11.console.scl3.mozilla.net
switch2.r202-11.console.scl3.mozilla.net
switch2.r601-11.console.scl3.mozilla.net
switch2.r602-11.console.scl3.mozilla.net

On the assumption that these will be marked as spare or decommed, I was going to remove them from nagios completely. Is there any reason I should not?

I didn't find any monitoring for the PDUs.

Comment 4

[Van Le [:van]]

Please remove them Nagios. The reason for the different hostnames is because one is for out of band access while the other is for inband access.  We recently added the inband hostnames because Nagios was giving false alerts on the .console/oob network.

Comment 5

[Amy Rich [:arr] [:arich]]

Removed with r92880