Jon,

Simply disable alerting on those ports. Problem solved. This is how you 
configure access switches that frequently have things plugged in / unplugged 
but you still want metrics when they are active.

On April 4, 2015 3:22:05 PM AKDT, Jon Auer <[email protected]> wrote:
>Fair enough, there are soup nazi aspects, though I've come to
>appreciate
>them after I got past my initial surprise. Ports that are enabled but
>unused being in alarm was the big one for me.
>
>Also, it does not require reverse DNS, only forward. I suspect syslog
>collection needs reverse DNS to match up with hosts but I don't use
>that...
>
>There are good non-soup-nazi reasons to require hostnames as
>identifiers
>instead of IPs, not the least of which is if you're using IPs now you
>make
>IPv6 compatability a problem. I could go on about the many benefits of
>hostname as device ID as opposed to IP or integer but I'm not trying to
>sell anything  :)
>
>
>On Sat, Apr 4, 2015 at 6:07 PM, Ken Hohhof <[email protected]> wrote:
>
>>   I stopped caring about Observium when someone said it wouldn’t
>monitor
>> something by IP address, it required both forward and reverse DNS. 
>Even
>> though Mike tried to defend that.  It’s “Soup Nazi” logic. 
>Everything else
>> in the world accepts an IP address in place of a hostname, I suspect
>you
>> have to write extra code to NOT do that.
>>
>>
>>  *From:* Jon Auer <[email protected]>
>> *Sent:* Saturday, April 04, 2015 5:52 PM
>> *To:* Animal Farm <[email protected]>
>> *Subject:* Re: [AFMUG] What Adam Armstrong of Observium thinks of
>WISPS
>>
>>  Until I see evidence to the contrary (just did a svn up and no wisp
>gear
>> has been removed...) I'm treating that entire exchange as the
>internet
>> equivalent of some drunk (Smeg) walking up to you(Adam) in the bar
>and
>> grabbing your arm and saying have your wife dance with me. You're
>like,
>> that's her choice and she says no. Repeating as the night goes on.
>> Eventually maybe you snap, say some unkind things because you want
>the
>> drunk to get lost and he just isn't taking the hint.
>>
>> That wasn't the first IRC exchange between them and others. Adam's
>> volatility is well known but in this case I believe he was sorely
>provoked.
>> We've all experienced the client from hell. Thing is, maybe some of
>us are
>> the client from hell.
>>
>> What happened in IRC after may help understand where he is coming
>from...
>>  [2015-04-04T01:39:59-0500] <BenA> [09:18:33] While I completely
>> understand that you're the author and shit, and what you say goes.
>> period... wasn't that a bit of an overreaction?
>> [2015-04-04T01:39:59-0500] <@adama> [09:18:56] BenA: you have no
>fucking
>> idea
>> [2015-04-04T01:39:59-0500] <BenA> [09:19:05] I don't, it's true.
>> [2015-04-04T01:39:59-0500] <@adama> [09:19:06] these people will not
>take
>> no for an answer
>> [2015-04-04T01:39:59-0500] <@adama> [09:19:07] they go on
>> [2015-04-04T01:39:59-0500] <@adama> [09:19:08] and on
>> [2015-04-04T01:39:59-0500] <@adama> [09:19:09] and on
>> [2015-04-04T01:39:59-0500] <@adama> [09:19:10] and on
>> [2015-04-04T01:39:59-0500] <BenA> [09:19:20] Couldn't you just, like,
>look
>> away from the screen?
>> [2015-04-04T01:39:59-0500] <@adama> [09:19:23] BUT MUH SUPPORT,
>MILLIONS
>> OF USERS, PLZ
>> [2015-04-04T01:39:59-0500] <@adama> [09:19:32] couldn't you?
>> [2015-04-04T01:39:59-0500] <BenA> [09:19:38] Ask him if he wants to
>pay.
>> if he doesn't, ignore?
>> [2015-04-04T01:39:59-0500] <@adama> [09:19:46] *oh, they all say they
>> want to pay*
>> [2015-04-04T01:39:59-0500] <BenA> [09:19:49] Ah.
>> [2015-04-04T01:39:59-0500] <@adama> [09:20:10] *but when they realise
>it
>> costs more than $5...*
>> [2015-04-04T01:39:59-0500] <BenA> [09:20:15] Right.
>> [2015-04-04T01:39:59-0500] <@adama> [09:20:31] the wireless industry
>seems
>> to be rammed full of overly entitled douches
>> [2015-04-04T01:39:59-0500] <@adama> [09:20:48] whot hink it's totally
>> justified that we support evey bit of shitty kit they have
>> [2015-04-04T01:39:59-0500] <@adama> [09:20:51] fucking wireless
>vendors
>> [2015-04-04T01:39:59-0500] <@adama> [09:21:00] every new MODEL they
>> release has to come with an entirely new set of mibs
>> [2015-04-04T01:39:59-0500] <@adama> [09:21:04] it's ridiculous
>> [2015-04-04T01:39:59-0500] <BenA> [09:21:11] You're a victim of your
>own
>> success.
>> [2015-04-04T01:39:59-0500] <@adama> [09:21:23] and none of these
>people
>> will ever listen when you tell them
>> [2015-04-04T01:39:59-0500] <@adama> [09:21:32] adn they just keep
>coming
>> back and coming back
>> [2015-04-04T01:39:59-0500] <@adama> [09:21:35] *he's asked here a few
>> times*
>> [2015-04-04T01:39:59-0500] <@adama> [09:21:40] *and we've had these
>> conversations on the mailing list*
>> [2015-04-04T01:39:59-0500] <@adama> [09:21:44] *and god knows what
>else*
>> [2015-04-04T01:39:59-0500] <@adama> [09:22:02] and you'd think
>> [2015-04-04T01:39:59-0500] <@adama> [09:22:03] when someone says
>> [2015-04-04T01:39:59-0500] <@adama> [09:22:07] "fuck off, it's not
>going
>> to happen"
>> [2015-04-04T01:39:59-0500] <@adama> [09:22:13] you'd realise that
>was,
>> well, that
>> [2015-04-04T01:39:59-0500] <BenA> [09:22:53] I guess he really sees
>the
>> advantages of Observium, and is keen for you to have further market
>> penetration... over your express wishes (-:
>> [2015-04-04T01:39:59-0500] <@adama> [09:23:06] no, he's keen to have
>his
>> own pet hardware supported
>> [2015-04-04T01:39:59-0500] <BenA> [09:23:12] That too.
>> [2015-04-04T01:39:59-0500] <@adama> [09:23:21] we already support 4
>> different families of cambium kit
>> [2015-04-04T01:39:59-0500] <@adama> [09:23:28] but apparently that
>doesn't
>> include the devices he has
>> [2015-04-04T01:39:59-0500] <@adama> [09:23:44] and i'm sure if i
>checked,
>> we'd only have like 2-3 users using cambium kit
>> [2015-04-04T01:39:59-0500] <BenA> [09:24:52] You could spend a few
>dozen
>> man months of eye-straining, back-breaking labour adding in some kind
>of
>> horribly complex and fucked up extensible API for 3rd party MIBs, and
>then
>> telling people they can add whatever the fuck support/devices they
>like,
>> but there's no support for it.
>> [2015-04-04T01:39:59-0500] <@adama> [09:25:06] or not
>> [2015-04-04T01:39:59-0500] <BenA> [09:25:14] Party-pooper.
>> [2015-04-04T01:39:59-0500] <BenA> [09:25:39] And people said you were
>a
>> nice guy.
>> [2015-04-04T01:39:59-0500] <@adama> [09:25:46] rarely
>> [2015-04-04T01:39:59-0500] <@adama> [09:25:55] and even more rarely
>in
>> here \o/
>> [2015-04-04T01:39:59-0500] <@adama> [09:26:49] *we have 3 cambium
>ptp400
>> devices and 5 cambium ptp800 devices*
>> [2015-04-04T01:39:59-0500] <@adama> [09:27:00] *in our entire
>userbase*
>> [2015-04-04T01:39:59-0500] <@adama> [09:27:24] *1086 cambium canopy
>> devices*
>> [2015-04-04T01:39:59-0500] <@adama> [09:27:31] *but, ofc, he said
>they
>> weren't important and not to bother with them*
>>
>>
>> On Sat, Apr 4, 2015 at 7:46 AM, Mike Hammett <[email protected]>
>wrote:
>>
>>>  I had got the impression that he didn't even want user
>contributions.
>>>
>>>
>>>
>>> -----
>>> Mike Hammett
>>> Intelligent Computing Solutions
>>> http://www.ics-il.com
>>>
>>> ------------------------------
>>> *From: *"Jon Auer" <[email protected]>
>>> *To: *"Animal Farm" <[email protected]>
>>> *Sent: *Saturday, April 4, 2015 2:59:53 AM
>>> *Subject: *Re: [AFMUG] What Adam Armstrong of Observium thinks of
>WISPS
>>>
>>> It's more than just OIDs, adding device support involves a fair
>amount of
>>> fiddly little things. Finding/cropping icon, regex to match the
>OS/device
>>> type to handle it correctly, logic to handle the device-specific
>things,
>>> logic to work around whatever they broke in the MIB (remember when
>Cambium
>>> returned strings instead of ints for some counters?). Then more
>testing.
>>>
>>> That's what makes Observium more useful out of the box than
>something
>>> like Cacti where you're adding OIDs onesey twosey to device
>templates.
>>>
>>>  I think a big part of his reaction is, if you watch IRC, for the
>past
>>> few months to years there have been people asking for WISP features
>and
>>> pretty much nobody in a place to write code to do it. My guess is he
>is
>>> time constrained and would rather work on other things (hence
>>> non-responsiveness to offers of money) combined with not wanting to
>deal
>>> with what could be perceived as self-entitled communication from
>some users.
>>>
>>> The hostile reaction to WISP gear:
>>>  CMMMicro is a switch that doesn't even use the switch MIB -> Work
>done
>>> to support WISP devices doesn't pay off in helping support other
>>> Enterprise/Wireline devices.
>>>
>>> Cambium is extra special because they version the PMP MIB against OS
>rev
>>> instead of starting out with a well-designed MIB as spec and fixing
>OS to
>>> match. The easy way out is to ignore that and use the latest but
>what
>>> happens when Cambium updates something? Bug reports from users on
>new OS
>>> complaining that something doesn't work. You update and now there's
>bug
>>> reports from the users that want to stay on old OS for a while. The
>hard
>>> way? Handle every OS rev differently/code gardening responsibility?
>You
>>> just can't win.
>>>
>>> <I digress>
>>> So, WISP gear, he doesn't need it and doesn't care. I need it and
>care so
>>> I write what I need. I may not  appreciate the politics of Observium
>but
>>> I'm being pragmatic. I contributed what little Cambium PMP device
>support
>>> there is in Observium currently and I have more devices I'd like to
>see
>>> supported. If the time comes that my contributions are turned away
>I'll
>>> look for another monitoring solution, not out of spite but because I
>need
>>> to monitor all the things.
>>>
>>> There may come a time when I move to LibreNMS. They seem to have
>openness
>>> & saying yes down but I want to see how they handle saying no to
>extraneous
>>> things/feature creep beyond monitoring metrics (e.g. if it were me,
>>> allow/keep rancid integration but just say no to generalized IPAM).
>>> You can't please everyone and who/how they choose to please will be
>>> insightful.
>>> </I digress>
>>>
>>> On Tue, Mar 31, 2015 at 5:06 PM, Mike Hammett <[email protected]>
>wrote:
>>>
>>>>  Do we know why Adam blows up whenever people specify OIDs they
>want to
>>>> track? I've never bothered to figure it out myself. He made it seem
>like
>>>> hte OID was such a small part of everything that needed to be done.
>>>>
>>>>
>>>>
>>>> -----
>>>> Mike Hammett
>>>> Intelligent Computing Solutions
>>>> http://www.ics-il.com
>>>>
>>>> ------------------------------
>>>>  *From: *"Neil Lathwood" <[email protected]>
>>>> *To: *[email protected]
>>>> *Sent: *Tuesday, March 31, 2015 1:08:23 PM
>>>> *Subject: *Re: [AFMUG] What Adam Armstrong of Observium thinks of
>WISPS
>>>>
>>>>  On 31 March 2015 at 19:04, WaveDirect <[email protected]>
>wrote:
>>>>
>>>>> Yeah you should accept at least equipment donations :)  Some of us
>may
>>>>> have spares we can part with and after you are done sell them to
>help buy
>>>>> other products you want to support.
>>>>>
>>>>
>>>> The donation of equipment is a huge ++++. It wouldn't be necessary
>to
>>>> send the kit anywhere just provide snmp access, that way we can see
>what
>>>> data is available and work on adding support.
>>>>
>>>> Thanks,
>>>>
>>>> Neil
>>>>
>>>>
>>>
>>>
>>>
>>
>>

-- 
Sent from my Android device with K-9 Mail. Please excuse my brevity.

Reply via email to