Network pics thread

Speaking of code violations, I always get a kick out of people who put their panel in a bathroom:
What section of the NEC is violated by this installation? Looks like the service panel is in a dry space -- inside the cubby.

Later: Oh, found it. 230.70(A)(2)
 
Last edited:
My point is, it's all plain and clear there in the picture. It's done wrong! Fire hazard no matter what. !

You don't use a 125V plug on a 230v circuit.

if yuo look at the picture closer, it's not just one conductor its 2 that is melted, so it had to have pulled a decent amount of current before it started to get warm..

Of course two connectors failed - line and neutral.

Kirchoff's current law.

Without seeing the inside of the plug, it's pointless to speculate on what the root cause is, because simply looking at the cable eliminates the possibility that it was shoddy assembly that burned it up. Alternatively, there could have been a lightning strike or the like.

I would argue that the most likely root cause was the scenario suggested by mikeblas, which is corroborated by evidence from Blue Fox - the entire inside of the plug was melted. Sustained overcurrent of 5A simply would not be sufficient to produce this kind of heating necessary. A quick look at a current handling table as a function of wire AWG shows that the 25A ampacity rating for #12AWG wire is taken assuming the conductor surface can be no hotter than 75 degrees C. Since that particular cable jacket is rated for 105 degrees C, we use the ampacity rating for the closest temperature figure, which is actually 90 degrees C (leaving a 15 degree margin of safety) - that ampacity rating is 30 amps.

If we look at the worst case current flow (30 amps sustained), and again - using the NEC's recommendation that only 80% of a circuit's capacity be used - that would only result in a sustained load of 24A - well within the ampacity rating of the cable, we can expect that the failure was not due to a sustained overcurrent, but rather due to arcing caused by poor assembly of the plug itself. Even at 30A running 24/7, the plug components (assuming the wire/plug interface has negligible resistance) do not have anything close to enough resistance to cause any thermal-related failure (P = (I^2) * R ) where R is measured in milliohms.

Of course the plug had an immense amount of current running through the two conductors that were damaged - note how they are line and neutral (the ground has that extra little tab). The most likely method of failure would be poor termination of a stranded-wire plug. The installer likely rushed and didn't see a few extra strands coming out of both L and N. Considering the amount of wire that's often stripped to terminate with these plugs, it's reasonably based upon the distance between the L and N screw that strands could have made contact - furthermore, even if the plug was never moved at all, thermal expansion and contraction could have pushed the two strands within sufficient proximity to arc.

Once the arc was struck, the current would drastically increase, explaining the failure of both L and N prongs. Considering we are only dealing with a couple strands out of a #12 wire, we can model these as a wire of around #20 gauge. Assuming that the circuit breaker acted extremely quickly (well under 500ms) we can look at the fusing current value and find it would take 834 amps to fuse #20 copper wire in about 32 milliseconds. Overcurrent-induced explosion in small-gauge copper wire is a totally plausible phenomenon, (this method of action has been exploited in some hobbyist high voltage projects), and would explain the severe damage to the inside of the plug.

Regarding the UPS selection - putting a 2200VA ups on a 30A circuit would be best practice. Considering the quality of PFC correction no most power supplies (where PF > 0.9), the 2200VA UPS would be moving nearly 2000W of real power. If we look at the NEC's recommendations for current use on circuits:

20A * 80% = 16A * 120V = 1920W
30A * 80% = 24A * 120V = 2880W

Using only a 20A circuit would likely run over 80% utilization (since the servers are not often power cycled, breaking this rule isn't the end of the world since it's designed to allow extra capacity for starting motors and the like - furthermore, the UPS would correct for voltage sags), but a 30A circuit would allow for plenty of headroom (making the NEC happy, which makes potential insurance companies happy).
 
Of course two connectors failed - line and neutral.

Kirchoff's current law.
Kirchhoff tells us that the sum of the current flowing into a junction is the same as the sum of the current flowing out of a junction. There's three junctions in this plug. One between the ground prong and the ground wire; one between the neutral prong and the neutral wire; and one between the hot prong and the hot wire.

How do you apply Kirchhoff to predict the failure of both the line and neutral connectors? (To be honest, I only see one failed prong, anyway. The discoloration is centered around the neutral prong; the ground prong looks scratched but not damaged and so does the hot prong. (Unless I'm mirroring the layout in my head.) )

It is heat, not current, that discolored and baked the plastic on the plug. An equal amount of current can flow from the hot prong to the equipment, out of the equipment, and back to the neutral prong while there's more heat at the neutral prong than the hot prong because of a low-resistance open connection or arcing. Kirchhoff doesn't tell us anything about heat, just current (and voltage, in his other law.)

Without seeing the inside of the plug, it's pointless to speculate on what the root cause is, because simply looking at the cable eliminates the possibility that it was shoddy assembly that burned it up. Alternatively, there could have been a lightning strike or the like.
Or foreign matter inside the plug, or a loose connection outside the plug like in the receptacle, or ... It seems quite remarkable that someone is able to make a diagnosis from a single picture without any other information. Or, maybe it's just that they're able to make a guess and what's remarkable is how certain and assertive they are about that guess.
 
It's been a while.

Events in my life have recently rekindled my interest in networking and now that I have a job again (yay!), I can afford to play with new toys now. I'll take a few pics of the horrendous mess I inherited at work later this week, and the miniature Cisco lab I've started (don't get too excited, just a 1760-V to putz around with CCME and a few 2950's for now).

In the meantime, here's pics of my rack from 4 years ago. Yeah I know, kinda minimalistic, but I just had a few IBM xSeries boxes that I was ESX'ing with and I was between phases. :) I do still have my Neatpatch and ProCurve, though.

jqFnJ.jpg


mpLen.jpg


0UZbN.jpg


ol0c2.jpg
 
so i posted some photos from the start of our DC build MONTHS ago, its been slow going but we have some progress. I can has UPS

IMG_20120820_133612.jpg


IMG_20120820_105900.jpg


IMG_20120820_160341.jpg


These are Huge, like half the size of my hand
IMG_20120820_160425.jpg


IMG_20120820_161100.jpg

IMG_20120820_161100.jpg


Generator Changeover
IMG_20120821_115336.jpg


Much more on its way soon!
 
Now there's a name I have not seen in a long time! Hope all is well! Nice to see you around.

Hehe, thanks man. Thought of ya when I was refreshing my LinkedIn account and saw your name there -- I've more or less been around, although I've only poked my head in once every few months or so. I had a change in venue and my wife had a baby so haven't had much time to browse here. Kinda silly, but I now have a bit more time on my hands than when she was just born.

Working for AT&T sucked.
 
I am sure as bad as AT&T is to deal with from the customer side, they are 10X worse to work for!

Not to get too off-topic, but I was a telephone guy. Well ok, I was actually a prem tech, but they do get telephone training. FIgured it was a job until I was able to find something else -- worst job I've ever had. I actually quit before I even had my current job. Oh, and unions suck, especially when they're as useless as the one I was forced to pay dues for.
 
Speaking of AT&T and PRI
lightspan.jpg


All that is left on it is the PRI, and a T1. The PRI is going away by the end of the week and hopefully the T1 is going away in a few months. Then we can unplugs this thing and push it down the street!
 
Ive seen tons of electrical short's the picture you have shown us is a plug that has PULLED way to much current, thus causing the weakest point to get hot and melt..

2200Va ups at 120v where ?

I got a APC Smart UPS 3000va at 120v.
 
We had 2 of those, they all needed to be set on fire. :)

Kidding aside, they're not bad, but not all that exciting.

Programming the phones and extensions for those Norstar systens is an irritating task. VoIP has really spoiled me in the last 7 or so years I've dealt with VoIP systems!
 
I dunno, it takes me longer to set up a phone on our Cisco Call Manager at work then it does on any of my phone systems at home...lol
 
So nothing fancy, but we just retired this fella yesterday. Circa 1999-2000 hardware, still running Windows NT 4.0 Server. Well one P2V later we can safely retire this thing and not worry about the hardware giving out

E3CDD6FD-7E68-4457-94C2-E6BC8F7439A3-10195-0000108F1181E173.jpg


17C3C510-3FBC-4E5F-BC09-7DB689F5681A-10195-0000108F1866BF33.jpg
 
^ Back when shit was made to last.

No kidding.. No raid just two drives partitioned out.. I think it would have lasted but still makes a guy nervous.. I accidentally knocked the power out the other day and wasn't sure it was going to start and some how it did....
 
No kidding.. No raid just two drives partitioned out.. I think it would have lasted but still makes a guy nervous.. I accidentally knocked the power out the other day and wasn't sure it was going to start and some how it did....

Hah hah hah. We had an NT4 box that was a file share for our Seibel installation (don't ask). There was a few million dollars of Seibel on top of another half a million in hardware and all of it was dependent on this ancient NT4 box that for various reasons couldn't be touched.

"Don't touch. Don't point. Don't even look at it."

http://www.youtube.com/watch?v=zZhrbE2rS1M

We never patched it. Never rebooted it. We were afraid to even look at the thing less we jinx it's amazing uptime.
 
Hah hah hah. We had an NT4 box that was a file share for our Seibel installation (don't ask). There was a few million dollars of Seibel on top of another half a million in hardware and all of it was dependent on this ancient NT4 box that for various reasons couldn't be touched.

"Don't touch. Don't point. Don't even look at it."

http://www.youtube.com/watch?v=zZhrbE2rS1M

We never patched it. Never rebooted it. We were afraid to even look at the thing less we jinx it's amazing uptime.

Same here... Cold boot converter cd cured hardware scares
 
I can't find a good cold boot converter image anywhere... I have some NT2K boxes that dreadfully need P2V'd. Any links?

lol I must be getting tired because I was like what the hell is P2V, then 5 mins later, I had a facepalm moment... Physical to Virtual right?
 
No kidding.. No raid just two drives partitioned out.. I think it would have lasted but still makes a guy nervous.. I accidentally knocked the power out the other day and wasn't sure it was going to start and some how it did....
those are the moments when you think you are gonna have a heart attack and then decide on a change :-P
 
I can't find a good cold boot converter image anywhere... I have some NT2K boxes that dreadfully need P2V'd. Any links?

I don't have any links. I had a buddy who had them for me and wasn't going to catch up to him as quick as I wanted too. So I scoured the internet looking for them to no avail.

lol I must be getting tired because I was like what the hell is P2V, then 5 mins later, I had a facepalm moment... Physical to Virtual right?

av-230759.gif
 
a pic of our new networking room(new core room) at a new school we setup. Nothing to special going on. Will be moving a bunch of servers to the room eventually and all of our fiber for all of the schools in the district comes back to this room,

corerack.jpg
 
Back
Top