Network pics thread

Skud · Feb 13, 2011

Mikeblas, thanks for the informative post.

mikeblas said:
A lot of the wiring shown here doesn't meet electrical code in the US. For example:

Section 110.26 of the NEC tells us that access to the service box must be unobstructed. That 2x2 is in the way, and it looks like there's plenty of other clutter defeating this requirement.

It's how the picture is.. That 2x2 is about 2' from the box. I can stand in front of it.. It still may not be "code" though. I'm not sure what the CEC code is. I know in the picture it looks like it's right up against the box.

mikeblas said:
Section E3305.3 of the building code says that plumbing can't be run above electrical panels. This gives an exception about plumbing above the structural ceiling, but it's clear that all the copper here is actually below the joists.

The code (NEC anyways) states that the panelboard area extends from the floor to ceiling for the depth and width of the box. While the plumbing is above, it's also "in front" of the panel box.

mikeblas said:
The "Neat and workmanlike" clause is found in Article 110, Section 12 of the NEC. It's unspecific and subject to lots of debate, but I think it's pretty clear this installation doesn't meet the intention of the requirement.

I'll agree with you on this one.. It's a bit of a spaghetti mess when all the wiring comes to the panel box. After ~4ft or so everything gets a bit neater.

mikeblas said:
NM cable must be protected. Here, it looks like it isn't. Protection needs to be provided because the hole is less than 1.25 inches from the face of the stud, leaving it unprotected. As installed here, a screw or nail through the drywall, then into the stud can penetrate the cable.

I actually just measured this and they're all through the center of the 2x4 and all are at least 1.5" from the front. One thing about the code though is that it only applies to "concealed" installations. The front of the wall is finished, the backside is an unfinished space (actually, a closet), so I don't know if this counts as concealed or not. Either way, it's over 1.25" from the face of the stud.

mikeblas said:
NM cable must be secured within 12 inches of every junction box, as specified in section 334.3 of the NEC. I think there's a box capacity issue here, too.

All the NM cables are secured right before they go into the box. The one coming in from the bottom actually goes above the box, secured to the wall then back in. I'm not sure on the box fill, but it's deep box.

mikeblas said:
Bicycle hooks aren't UL-listed for in-wall installation to hold electrical wires; even limited energy communication wires. This is covered in NEC sections 800.24 and 300.4(D). Since cable TV wiring is involved, there are also relevant sections in Article 820.

I suppose this could be an issue in that these aren't to code, but will be super easy to replace if that becomes an issue.

mikeblas said:
NM cable must be secured at intervals not exceeding 4.5 feet. Close call, but the picture looks like there's only one staple for a pretty long run. This is NEC 334.3 again.

Again this is because of the camera shot. I'm sure if I walked around a building looking through a camera lens I could find all sorts of "code violations". The wire is secured about every 2.5ft.

mikeblas said:
I see that Skud's pictures are hosted at a .CA site, so maybe the installation is in Canada. Most of the same requirements apply, and the relevant sections of the CSA should be referenced.

Yes, I'm in Canada. I'm not a house inspector, but I can say we had the house inspected before the purchase and it met all codes at the time of construction.

Riley

mikeblas · Feb 13, 2011

Skud said:
Yes, I'm in Canada. I'm not a house inspector, but I can say we had the house inspected before the purchase and it met all codes at the time of construction.

Riley

Yes, it's sometimes hard to tell distances from pictures because of parallax.

How do Canadian home inspectors go about verifying all construction, wiring, plumbing, and mechanical is up to appropriate code? Such an inspection must take days!

Skud · Feb 13, 2011

mikeblas said:
Yes, it's sometimes hard to tell distances from pictures because of parallax.

How do Canadian home inspectors go about verifying all construction, wiring, plumbing, and mechanical is up to appropriate code? Such an inspection must take days!

I should be more specific and say we had a "home" (general) inspection and electrical inspection. Took about a 8 hours for each.. The general home inspector was unfamiliar with the boiler/water heating system so he brought in a licensed inspector for that portion too.

Riley

jeffmoss26 · Feb 13, 2011

Mike, in regards to the first picture, what do you mean by the cables being suspended?

mikeblas · Feb 13, 2011

At each floor, the cable needs to be clamped so the weight of the cable is transferred to the floor, not just hanging.

Red Squirrel · Feb 13, 2011

Hmm that makes sense, if you had a building say, 10 floors, that would be a lot of weight on the clamps at the top. Hard to clamp that many cables without squashing them though, not sure what would be the proper way to do it.

I think it is legal in a case where the wires are fished, so there is no way to clamp them, but I may be wrong on this.

Red Squirrel · Feb 13, 2011

Jesse B said:
I may or may not know a guy

I love your setup. You should run some benchmarks on it for us

As requested.

My main server (5x1TB local mdadm raid 5 with Hitachi 7k1000 drives) :

Code:

[root@borg tmp]# iozone -s 100m -t 4 -F a b c d
	Iozone: Performance Test of File I/O
	        Version $Revision: 3.239 $
		Compiled for 64 bit mode.
		Build: linux 

	Contributors:William Norcott, Don Capps, Isom Crawford, Kirby Collins
	             Al Slater, Scott Rhine, Mike Wisner, Ken Goss
	             Steve Landherr, Brad Smith, Mark Kelly, Dr. Alain CYR,
	             Randy Dunlap, Mark Montague, Dan Million, 
	             Jean-Marc Zucconi, Jeff Blomberg,
	             Erik Habbinga, Kris Strecker, Walter Wong.

	Run began: Sun Feb 13 17:57:05 2011

	File size set to 102400 KB
	Command line used: iozone -s 100m -t 4 -F a b c d
	Output is in Kbytes/sec
	Time Resolution = 0.000001 seconds.
	Processor cache size set to 1024 Kbytes.
	Processor cache line size set to 32 bytes.
	File stride size set to 17 * record size.
	Throughput test with 4 processes
	Each process writes a 102400 Kbyte file in 4 Kbyte records

	Children see throughput for  4 initial writers 	=  248242.31 KB/sec
	Parent sees throughput for  4 initial writers 	=  122422.84 KB/sec
	Min throughput per process 			=   48562.48 KB/sec 
	Max throughput per process 			=   84471.23 KB/sec
	Avg throughput per process 			=   62060.58 KB/sec
	Min xfer 					=   58896.00 KB

	Children see throughput for  4 rewriters 	=  237088.15 KB/sec
	Parent sees throughput for  4 rewriters 	=  139848.59 KB/sec
	Min throughput per process 			=   53288.08 KB/sec 
	Max throughput per process 			=   71955.09 KB/sec
	Avg throughput per process 			=   59272.04 KB/sec
	Min xfer 					=   81968.00 KB

	Children see throughput for  4 readers 		= 4598808.56 KB/sec
	Parent sees throughput for  4 readers 		= 2168206.24 KB/sec
	Min throughput per process 			= 1020941.06 KB/sec 
	Max throughput per process 			= 1488306.62 KB/sec
	Avg throughput per process 			= 1149702.14 KB/sec
	Min xfer 					=   57820.00 KB

	Children see throughput for 4 re-readers 	= 4063285.00 KB/sec
	Parent sees throughput for 4 re-readers 	= 3621618.73 KB/sec
	Min throughput per process 			=  897562.81 KB/sec 
	Max throughput per process 			= 1242987.62 KB/sec
	Avg throughput per process 			= 1015821.25 KB/sec
	Min xfer 					=   57812.00 KB

	Children see throughput for 4 reverse readers 	= 1591692.88 KB/sec
	Parent sees throughput for 4 reverse readers 	= 1032894.53 KB/sec
	Min throughput per process 			=     282.58 KB/sec 
	Max throughput per process 			= 1589987.75 KB/sec
	Avg throughput per process 			=  397923.22 KB/sec
	Min xfer 					=      24.00 KB

	Children see throughput for 4 stride readers 	= 1892889.44 KB/sec
	Parent sees throughput for 4 stride readers 	= 1745339.44 KB/sec
	Min throughput per process 			=   77439.15 KB/sec 
	Max throughput per process 			= 1649241.12 KB/sec
	Avg throughput per process 			=  473222.36 KB/sec
	Min xfer 					=    4852.00 KB

	Children see throughput for 4 random readers 	= 1959287.93 KB/sec
	Parent sees throughput for 4 random readers 	= 1780546.46 KB/sec
	Min throughput per process 			=    1058.31 KB/sec 
	Max throughput per process 			= 1952179.62 KB/sec
	Avg throughput per process 			=  489821.98 KB/sec
	Min xfer 					=      60.00 KB

	Children see throughput for 4 mixed workload 	= 2529798.64 KB/sec
	Parent sees throughput for 4 mixed workload 	=   14807.28 KB/sec
	Min throughput per process 			=    3088.89 KB/sec 
	Max throughput per process 			= 1404778.50 KB/sec
	Avg throughput per process 			=  632449.66 KB/sec
	Min xfer 					=     228.00 KB

	Children see throughput for 4 random writers 	=    5760.11 KB/sec
	Parent sees throughput for 4 random writers 	=    4468.75 KB/sec
	Min throughput per process 			=    1118.94 KB/sec 
	Max throughput per process 			=    1724.20 KB/sec
	Avg throughput per process 			=    1440.03 KB/sec
	Min xfer 					=   69772.00 KB

	Children see throughput for 4 pwrite writers 	=  158407.49 KB/sec
	Parent sees throughput for 4 pwrite writers 	=   83174.60 KB/sec
	Min throughput per process 			=   36864.23 KB/sec 
	Max throughput per process 			=   44441.26 KB/sec
	Avg throughput per process 			=   39601.87 KB/sec
	Min xfer 					=   84768.00 KB

	Children see throughput for 4 pread readers 	= 5418532.62 KB/sec
	Parent sees throughput for 4 pread readers 	=  477320.34 KB/sec
	Min throughput per process 			= 1049409.62 KB/sec 
	Max throughput per process 			= 1566392.62 KB/sec
	Avg throughput per process 			= 1354633.16 KB/sec
	Min xfer 					=   68348.00 KB



iozone test complete.

(done while VMs and other stuff going on said array, so it may have offset stats a bit)

1st enclosure (hooked straight to SuperMicro Atom box via FC) :

Code:

[root@borg vg1a]# iozone -s 100m -t 4 -F a b c d
	Iozone: Performance Test of File I/O
	        Version $Revision: 3.239 $
		Compiled for 64 bit mode.
		Build: linux 

	Contributors:William Norcott, Don Capps, Isom Crawford, Kirby Collins
	             Al Slater, Scott Rhine, Mike Wisner, Ken Goss
	             Steve Landherr, Brad Smith, Mark Kelly, Dr. Alain CYR,
	             Randy Dunlap, Mark Montague, Dan Million, 
	             Jean-Marc Zucconi, Jeff Blomberg,
	             Erik Habbinga, Kris Strecker, Walter Wong.

	Run began: Sun Feb 13 18:21:24 2011

	File size set to 102400 KB
	Command line used: iozone -s 100m -t 4 -F a b c d
	Output is in Kbytes/sec
	Time Resolution = 0.000001 seconds.
	Processor cache size set to 1024 Kbytes.
	Processor cache line size set to 32 bytes.
	File stride size set to 17 * record size.
	Throughput test with 4 processes
	Each process writes a 102400 Kbyte file in 4 Kbyte records

	Children see throughput for  4 initial writers 	=  111005.44 KB/sec
	Parent sees throughput for  4 initial writers 	=   43575.32 KB/sec
	Min throughput per process 			=   20878.63 KB/sec 
	Max throughput per process 			=   36790.57 KB/sec
	Avg throughput per process 			=   27751.36 KB/sec
	Min xfer 					=   63872.00 KB

	Children see throughput for  4 rewriters 	=  143076.69 KB/sec
	Parent sees throughput for  4 rewriters 	=   51869.24 KB/sec
	Min throughput per process 			=   13113.60 KB/sec 
	Max throughput per process 			=   64435.82 KB/sec
	Avg throughput per process 			=   35769.17 KB/sec
	Min xfer 					=   32096.00 KB

	Children see throughput for  4 readers 		= 5261332.50 KB/sec
	Parent sees throughput for  4 readers 		= 2433673.83 KB/sec
	Min throughput per process 			=  957725.75 KB/sec 
	Max throughput per process 			= 1832364.50 KB/sec
	Avg throughput per process 			= 1315333.12 KB/sec
	Min xfer 					=   25848.00 KB

	Children see throughput for 4 re-readers 	= 4182656.81 KB/sec
	Parent sees throughput for 4 re-readers 	= 4088726.99 KB/sec
	Min throughput per process 			=  606457.94 KB/sec 
	Max throughput per process 			= 1560332.12 KB/sec
	Avg throughput per process 			= 1045664.20 KB/sec
	Min xfer 					=   39820.00 KB

	Children see throughput for 4 reverse readers 	= 4458046.94 KB/sec
	Parent sees throughput for 4 reverse readers 	= 4408194.11 KB/sec
	Min throughput per process 			= 1039869.69 KB/sec 
	Max throughput per process 			= 1182148.38 KB/sec
	Avg throughput per process 			= 1114511.73 KB/sec
	Min xfer 					=   90708.00 KB

	Children see throughput for 4 stride readers 	= 4521530.50 KB/sec
	Parent sees throughput for 4 stride readers 	= 4307224.95 KB/sec
	Min throughput per process 			= 1009564.88 KB/sec 
	Max throughput per process 			= 1230519.88 KB/sec
	Avg throughput per process 			= 1130382.62 KB/sec
	Min xfer 					=   82656.00 KB

	Children see throughput for 4 random readers 	= 4804324.19 KB/sec
	Parent sees throughput for 4 random readers 	= 4316371.60 KB/sec
	Min throughput per process 			=  821574.44 KB/sec 
	Max throughput per process 			= 1364518.38 KB/sec
	Avg throughput per process 			= 1201081.05 KB/sec
	Min xfer 					=   32716.00 KB

	Children see throughput for 4 mixed workload 	= 3408043.19 KB/sec
	Parent sees throughput for 4 mixed workload 	=    3863.67 KB/sec
	Min throughput per process 			=  509435.94 KB/sec 
	Max throughput per process 			= 1243729.12 KB/sec
	Avg throughput per process 			=  852010.80 KB/sec
	Min xfer 					=   41816.00 KB

	Children see throughput for 4 random writers 	=    4072.62 KB/sec
	Parent sees throughput for 4 random writers 	=    2009.46 KB/sec
	Min throughput per process 			=     946.45 KB/sec 
	Max throughput per process 			=    1206.65 KB/sec
	Avg throughput per process 			=    1018.15 KB/sec
	Min xfer 					=   80492.00 KB

	Children see throughput for 4 pwrite writers 	=  125795.55 KB/sec
	Parent sees throughput for 4 pwrite writers 	=   40691.47 KB/sec
	Min throughput per process 			=   20441.10 KB/sec 
	Max throughput per process 			=   41609.36 KB/sec
	Avg throughput per process 			=   31448.89 KB/sec
	Min xfer 					=   57508.00 KB

	Children see throughput for 4 pread readers 	= 4860058.56 KB/sec
	Parent sees throughput for 4 pread readers 	= 2810412.37 KB/sec
	Min throughput per process 			=  719989.56 KB/sec 
	Max throughput per process 			= 1389247.50 KB/sec
	Avg throughput per process 			= 1215014.64 KB/sec
	Min xfer 					=   53092.00 KB



iozone test complete.

2nd enclosure (daisy chainned to 1st via FC) :

Code:

[root@borg vg0a]# iozone -s 100m -t 4 -F a b c d
	Iozone: Performance Test of File I/O
	        Version $Revision: 3.239 $
		Compiled for 64 bit mode.
		Build: linux 

	Contributors:William Norcott, Don Capps, Isom Crawford, Kirby Collins
	             Al Slater, Scott Rhine, Mike Wisner, Ken Goss
	             Steve Landherr, Brad Smith, Mark Kelly, Dr. Alain CYR,
	             Randy Dunlap, Mark Montague, Dan Million, 
	             Jean-Marc Zucconi, Jeff Blomberg,
	             Erik Habbinga, Kris Strecker, Walter Wong.

	Run began: Sun Feb 13 17:41:24 2011

	File size set to 102400 KB
	Command line used: iozone -s 100m -t 4 -F a b c d
	Output is in Kbytes/sec
	Time Resolution = 0.000001 seconds.
	Processor cache size set to 1024 Kbytes.
	Processor cache line size set to 32 bytes.
	File stride size set to 17 * record size.
	Throughput test with 4 processes
	Each process writes a 102400 Kbyte file in 4 Kbyte records

	Children see throughput for  4 initial writers 	=  104974.53 KB/sec
	Parent sees throughput for  4 initial writers 	=   48634.14 KB/sec
	Min throughput per process 			=   24714.75 KB/sec 
	Max throughput per process 			=   29823.42 KB/sec
	Avg throughput per process 			=   26243.63 KB/sec
	Min xfer 					=   87344.00 KB

	Children see throughput for  4 rewriters 	=  202151.63 KB/sec
	Parent sees throughput for  4 rewriters 	=   49410.94 KB/sec
	Min throughput per process 			=    9011.29 KB/sec 
	Max throughput per process 			=   94112.12 KB/sec
	Avg throughput per process 			=   50537.91 KB/sec
	Min xfer 					=   10784.00 KB

	Children see throughput for  4 readers 		= 5032917.12 KB/sec
	Parent sees throughput for  4 readers 		= 2953475.51 KB/sec
	Min throughput per process 			= 1161470.88 KB/sec 
	Max throughput per process 			= 1367432.88 KB/sec
	Avg throughput per process 			= 1258229.28 KB/sec
	Min xfer 					=   88964.00 KB

	Children see throughput for 4 re-readers 	= 4309418.00 KB/sec
	Parent sees throughput for 4 re-readers 	= 4244995.93 KB/sec
	Min throughput per process 			=  879202.00 KB/sec 
	Max throughput per process 			= 1262702.75 KB/sec
	Avg throughput per process 			= 1077354.50 KB/sec
	Min xfer 					=   71612.00 KB

	Children see throughput for 4 reverse readers 	= 4110882.69 KB/sec
	Parent sees throughput for 4 reverse readers 	= 3988275.68 KB/sec
	Min throughput per process 			=  637964.19 KB/sec 
	Max throughput per process 			= 1353658.75 KB/sec
	Avg throughput per process 			= 1027720.67 KB/sec
	Min xfer 					=   47640.00 KB

	Children see throughput for 4 stride readers 	= 3968449.91 KB/sec
	Parent sees throughput for 4 stride readers 	= 3857166.62 KB/sec
	Min throughput per process 			=  418668.22 KB/sec 
	Max throughput per process 			= 1545242.38 KB/sec
	Avg throughput per process 			=  992112.48 KB/sec
	Min xfer 					=   29048.00 KB

	Children see throughput for 4 random readers 	= 3864625.50 KB/sec
	Parent sees throughput for 4 random readers 	= 3650929.95 KB/sec
	Min throughput per process 			=  605767.81 KB/sec 
	Max throughput per process 			= 1339177.50 KB/sec
	Avg throughput per process 			=  966156.38 KB/sec
	Min xfer 					=   33032.00 KB

	Children see throughput for 4 mixed workload 	= 3160102.97 KB/sec
	Parent sees throughput for 4 mixed workload 	=    3550.99 KB/sec
	Min throughput per process 			=  448576.12 KB/sec 
	Max throughput per process 			= 1319211.38 KB/sec
	Avg throughput per process 			=  790025.74 KB/sec
	Min xfer 					=   33680.00 KB

	Children see throughput for 4 random writers 	=    4242.87 KB/sec
	Parent sees throughput for 4 random writers 	=    1681.49 KB/sec
	Min throughput per process 			=     762.09 KB/sec 
	Max throughput per process 			=    1384.98 KB/sec
	Avg throughput per process 			=    1060.72 KB/sec
	Min xfer 					=   56776.00 KB

	Children see throughput for 4 pwrite writers 	=  135172.08 KB/sec
	Parent sees throughput for 4 pwrite writers 	=   39707.31 KB/sec
	Min throughput per process 			=   22442.86 KB/sec 
	Max throughput per process 			=   48561.72 KB/sec
	Avg throughput per process 			=   33793.02 KB/sec
	Min xfer 					=   56660.00 KB

	Children see throughput for 4 pread readers 	= 5586045.19 KB/sec
	Parent sees throughput for 4 pread readers 	= 2465929.86 KB/sec
	Min throughput per process 			=  991338.56 KB/sec 
	Max throughput per process 			= 1656676.88 KB/sec
	Avg throughput per process 			= 1396511.30 KB/sec
	Min xfer 					=   55540.00 KB



iozone test complete.

Actually my first time running benchmarks on any of my systems so not really sure whether these are good or not, don't really have anything to compare to.

The 1U SuperMicro box is connected to my main network via gigabit and the iSCSI traffic is on said network as well. Not to mention I'm just using a home grade dlink router/switch. So the bottleneck is most likely the gigabit connection. Since that box has two nics, eventually I want to team them but I don't have a managed switch yet.

Oh and file system is ext3 on all devices.

I have 3 1TB hitachis that I pulled out of my main server due to lot of IO errors, not sure if they are failing or if I have other issues with the backplane itself, but either way, they still work, they just make dmesg talk a lot. Going to make a LUN to see if the enclosure takes bigger drives and run a separate benchmark on that as well. I might do a raid 0 just to live on the edge.

What I think I might do is leave one enclosure on all the time, as it does not use that much power on it's own, and the others will be for messing around/archiving and what not. I will probably backup my p2p downloads on it as those don't get backed up anywhere currently. It's stuff that is replaceable and takes up lot of space, but would still be a pita to rebuild if I did lose it. Come to think of it, I could probably just download my stuff directly to it, then organize it on my main server, so only keep the good stuff. I'm really enjoying playing with iSCSI, it's a very cool technology and surprisingly very easy to use in Linux and Windows.

Later on I'm going to play around with the McData switch, but I probably won't end up using it. It might be for trade.

Edit:

Unfortunatly, I don't think it accepts other drives. Either size is too big, or the firmware on the drive may have to be flashed to be "IBM Approved" so they can rape you when you want more space.

I get a bunch of these errors with my 1TB drive plugged in:

Code:

sd 4:0:12:0: [sdm] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE,SUGGEST_OK
sd 4:0:12:0: [sdm] Sense Key : Aborted Command [current]
sd 4:0:12:0: [sdm] <<vendor>> ASC=0x88 ASCQ=0x2ASC=0x88 ASCQ=0x2
end_request: I/O error, dev sdm, sector 0

While the drive itself is failing, it was still coming up ok in the server, here it wont come up at all. Doing it to all 3 drives too.

This kinda sucks though, as I was intending to add bigger drives over time. I'll have to do more research at some point to see if there is any way to bypass that, but it would probably be hard to find such info.

jeffmoss26 · Feb 13, 2011

Hard to say...that's a vertical riser. I would support them if possible. I'm guessing the whole thing will be enclosed in drywall though.

MikeTrike · Feb 13, 2011

Just finished cleaning up our server room.

Wiring runner.

Temporary cables dangling.

For these yet to be virtualized servers.

Red Squirrel · Feb 13, 2011

Nice! I like the idea of the cable runner. Getting power to the rack can be tricky sometimes as you normally want the feeder to be on the bottom. Been debating how I want to setup my rack, as currently it's a mess and is just temporary.

You should put S hooks on that cable and hang stuff off it, like spare patch cords. j/k

MikeTrike · Feb 13, 2011

Red Squirrel said:
Nice! I like the idea of the cable runner. Getting power to the rack can be tricky sometimes as you normally want the feeder to be on the bottom. Been debating how I want to setup my rack, as currently it's a mess and is just temporary.

You should put S hooks on that cable and hang stuff off it, like spare patch cords. j/k

Yeah unfortunately raised flooring was not an option, so that heavy duty garage cable runner was a life saver, it's heavy and rock solid, you can jump on it, kick it, etc and it does not move. It's all sorts of wedged between the rack and the wall.

I should also have those dangling cables gone by next weekend once those boxes are virtualized.

I was also going to rip out the carpet, but it's all sorts of glued down hardcore.

Captain Colonoscopy · Feb 13, 2011

I like that cable running thingy. What make/model is it? Might have to look at some of those for a few of my clients.

Red Squirrel · Feb 13, 2011

Ha never even noticed the carpet. Yeah carpet in a server room is usually not the greatest idea.

We recently got the carpet removed from our IT office. That was quite the ordeal. We have these HUGE L shaped desks, industrial shelving unit etc... glad it was not us that had to move all that but we did have to pack all our stuff in boxes. Glad it's over. Got vinyl tile floor to match the server room now.

The main reason was the carpet was very dirty from many years of spilled coffees, and it also produced lot of static.

MikeTrike · Feb 13, 2011

Captain Colonoscopy said:
I like that cable running thingy. What make/model is it? Might have to look at some of those for a few of my clients.

Got it from here:
http://cableorganizer.com/cord-covers/high-capacity-covers.htm

MikeTrike · Feb 13, 2011

Red Squirrel said:
The main reason was the carpet was very dirty from many years of spilled coffees, and it also produced lot of static.

Yeah, static is a bitch in my office and the server room. I had a dedicated and direct heavy duty ground wire anchored to the cabinets. So every time I go into the server room, I touch the cabinet and feel a nice big pop of static; just to be as safe as possible.

Captain Colonoscopy · Feb 14, 2011

MikeTrike said:
Got it from here:
http://cableorganizer.com/cord-covers/high-capacity-covers.htm

sweet, thanks!

andypress · Feb 14, 2011

Hallis said:
Awesome andypress:

I've got a 3U Rackable 12-bay dual opteron 270 system on the way to serve as a NAS. Hope it's not a gigantic piece of crap. Have you been suffering a lot of hardware failures with those PSU's? Should I avoid using one?

I don't think you should have any issues; the hardware in the 3u opteron 270 is more robust, not to mention a few years newer [than my racksys]. A year back I found a used systems like your's without processors (3u, 12-bay rackable}. The notable differences: It's power supply is 100w stronger. 12 modern sata drives should be fairly power efficient than ide. The components should have better cooling with the larger case, however it still dense.

As a side note, all 40 of the 2u Opteron 246 Rackable systems we have are nearly identical. They were probably manufactured in 2003, as they are all C0 revision processors. Microsoft used a color coded sticker on each machine; red on a drive tray indicates that the hard drive was replaced (as well as the date).

Specs:
2u Dual Socket 940 w/ two Opteron 246 (single core)
4 x 250gb ide, Maxtor (12w draw)
8gb ddr333
2 x 80mm @8000rpm case fans. Move about 250cfm each, loud as hell
1u heatsink for each proc. At idle, cpu temps are around 60C
350watt 1u power supply (about seven years old at this point)

I attribute the psu failures to two issues: seven year-aged capacitors, and heavy stress - current load and high heat.

Berg0 · Feb 14, 2011

MikeTrike said:
Yeah, static is a bitch in my office and the server room. I had a dedicated and direct heavy duty ground wire anchored to the cabinets. So every time I go into the server room, I touch the cabinet and feel a nice big pop of static; just to be as safe as possible.

That is scary. You can buy ESD floor mats, or lay ESD proof lino overtop of the carpet I beleive. Every server room i've worked in has been purpose built, so I guess I'm just used to not having to deal with that.

valve1138 · Feb 14, 2011

MikeTrike said:
Yeah, static is a bitch in my office and the server room. I had a dedicated and direct heavy duty ground wire anchored to the cabinets. So every time I go into the server room, I touch the cabinet and feel a nice big pop of static; just to be as safe as possible.

How dry is it in there? Sounds like you need a humidifier.

MikeTrike · Feb 14, 2011

Berg0 said:
That is scary. You can buy ESD floor mats, or lay ESD proof lino overtop of the carpet I beleive. Every server room i've worked in has been purpose built, so I guess I'm just used to not having to deal with that.

I just ordered some anti-static mats to run along the main walking path for the cabs.

valve1138 said:
How dry is it in there? Sounds like you need a humidifier.

It got dry when the weather got super cold, I'm going to try and pick up some monitoring gear that I can hook all of that up to. Any suggestions for something that is semi-automated?

valve1138 · Feb 14, 2011

You might be able to tie a humidifier in with the AC unit so it takes care of itself.

Adam · Feb 14, 2011

mikeblas said:
A lot of the wiring shown here doesn't meet electrical code in the US. For example:

Calbing passing between floors needs firestopping in and around the ports that penetrate structural members. The cabling also needs to be suspended at each floor. You can find more about this in sections 800.3(C) and 300.21 of the National Electric Code (NEC). Hopefully, these are all CMR- or CMP-rated ables, per section 800.179(B).

Thanks... we're not finished yet FYI... the whole place is still under construction.

ltickett · Feb 14, 2011

Now looking at OneStor/Xyratec 5220FA-220 2U Rack Mount 12-Bay 2 Channel Fibre Channel to SATA Disk Storage System as i know it can at least handle 500GB disks... i'm diving into the unknown a little bit now though!

3 more of the SuperMicro SuperServers arrived this morning

Berg0 · Feb 14, 2011

ltickett said:
Now looking at OneStor/Xyratec 5220FA-220 2U Rack Mount 12-Bay 2 Channel Fibre Channel to SATA Disk Storage System as i know it can at least handle 500GB disks... i'm diving into the unknown a little bit now though!

3 more of the SuperMicro SuperServers arrived this morning

xyratex is pretty common. NetApp uses them.

jeffmoss26 · Feb 15, 2011

We use those cable covers here too for electrical/sound stuff outside...http://www.yjams.com/

Jay_2 · Feb 16, 2011

Jesse B · Feb 16, 2011

What's that small 4"-ish screen on top of your cabinet? Nice looking setup as well

Jay_2 · Feb 16, 2011

Its an AlphaCool External LCD display plugged into my WHS via USB. It shows disk space, memory use, current bandwidth, current CPU use per core, temps etc etc and its all graphed on indiviual pages that it flicks between.

dashpuppy · Feb 16, 2011

Jay_oasis said:
Its an AlphaCool External LCD display plugged into my WHS via USB. It shows disk space, memory use, current bandwidth, current CPU use per core, temps etc etc and its all graphed on indiviual pages that it flicks between.

I like, i want

where by from ?

MikeTrike · Feb 16, 2011

dashpuppy said:
I like, i want where by from ?

You might be able to find it here

Jesse B · Feb 16, 2011

Jay_oasis said:
Its an AlphaCool External LCD display plugged into my WHS via USB. It shows disk space, memory use, current bandwidth, current CPU use per core, temps etc etc and its all graphed on indiviual pages that it flicks between.

That's seriously cool!

Jay_2 · Feb 16, 2011

I don't think they make them anymore. I have a 5.25" version in my PC as well. Love them!

=

Red Squirrel · Feb 16, 2011

I wish stuff like that was more readily available, this is really cool, and would be awesome for custom built servers too.

C7J0yc3 · Feb 16, 2011

Red Squirrel said:
I wish stuff like that was more readily available, this is really cool, and would be awesome for custom built servers too.

FrozenCPU and Performance PC's both still sell these screens. They are expensive though, last I checked they are around $100

mikeblas · Feb 16, 2011

Of course they make 'em. See http://www.matrixorbital.com/ , for example. They're not often used for servers because nobody's around to look at servers.

Jay_2 · Feb 17, 2011

I mean they don't make the AquaCool ones any more

mikeblas · Feb 17, 2011

Oh! Sorry.

dashpuppy · Feb 17, 2011

Well i recieved my new Dell 24port Gigabit switch AHHHH so nice to go from 10/100 to gigabit.

got a Sonicwall T210, that i use now that i like. ( free from work fully licensed.

yes it's messy, just waiting to buy a rack.

Got my new server running too! ( original board was defective.)

16gigs ecc ddr3 ram i3 processor 5 hdd's 2 raid sets 2 @ mirror 500's and 3 raided main os drives.

aaronearles · Feb 17, 2011

What are you running on the server, I'm just curious why you paired 16GB up with an i3?

Berg0 · Feb 17, 2011

aaronearles said:
What are you running on the server, I'm just curious why you paired 16GB up with an i3?

only logical thing I can think of would be a caching proxy

Network pics thread

Gawd

[H]ard|DCer of the Month - May 2006

Gawd

2[H]4U

[H]ard|DCer of the Month - May 2006

[H]F Junkie

[H]F Junkie

2[H]4U

Fully [H]

[H]F Junkie

Fully [H]

2[H]4U

[H]F Junkie

Fully [H]

Fully [H]

2[H]4U

n00b

[H]ard|Gawd

[H]F Junkie

Fully [H]

[H]F Junkie

[H]ard|Gawd

[H]ard|Gawd

[H]ard|Gawd

2[H]4U

2[H]4U

[H]ard|Gawd

2[H]4U

Supreme [H]ardness

Fully [H]

[H]ard|Gawd

2[H]4U

[H]F Junkie

[H]ard|Gawd

[H]ard|DCer of the Month - May 2006

2[H]4U

[H]ard|DCer of the Month - May 2006

Supreme [H]ardness

[H]ard|Gawd

[H]ard|Gawd