cable testing?
I have run a couple of cat5e cables and I am trying to terminate them, unsuccessfully at present. Will a cable tester help me? I suspect that each time I put a plug on the end one or more of the wires is in the wrong place, or not quite long enough to make the connection. Coupled with this I am only 90% sure which cable end is which at the switch end (ie the centre of the star), having failed to mark them. Is there some sort of cable tester that can, eg, tell me what wires are right and what are wrong, and which end of the cable is wired wrong? And, heres the hit, can someone in ChCh lend me one? Desperate and frustrated in Linwood. N
Re: Filesystem and replacing .. The final word??
Peter Glassenbury (CSSE) wrote: (vi works all the time :-) ) That is one of vi's biggest advantages over Emacs, but then Emacs can act as much more than just a text editor. I still want to learn how to use vi efficiently. ...some day. :-) --Aidan Gauland signature.asc Description: OpenPGP digital signature
Re: cable testing?
I have run a couple of cat5e cables and I am trying to terminate them, unsuccessfully at present. Will a cable tester help me? Depends on the problem, but prob yes. If you broke copper core pulling the cables the tester should show that (not fix it though). If you mixed up cables it'll show that too - but only after the crimping. I suspect that each time I put a plug on the end one or more of the wires is in the wrong place, or not quite long enough to make the connection. Only improved technique will help there. The tester will only show a bad job afterwards. Coupled with this I am only 90% sure which cable end is which at the switch end (ie the centre of the star), having failed to mark them. That's the easier bit, assuming you have easy access to all 8 ends of both cat5e cables on both sides, and you are not too colour blind to tell the colouring apart. Connect the ends of a pair of one cable together on one side. On the other side, use a multimeter (ohm setting) to see which 2 ends of the same coloured pair are shortened - that's the same cable. There are a gazillion ways to do similar things, like use a plugpack/battery and a torch bulb. Specialised equipment made for this purpose is lacking the geek factor... Is there some sort of cable tester that can, eg, tell me what wires are right and what are wrong, and which end of the cable is wired wrong? Yes, basic cheap ones run a continuity test on all 8 wires. They tell if some are swapped, but NOT if the 4 pairs are kept paired (i.e. if there is a double-fault which cancels itself), in which case it'll show ok for a cable which doesn't work. You MUST keep the pairs paired, though you can swap one pair with another as long as you do it the same on both ends. (I strongly recommend you stick with the prescribed colour scheme, or you're risking your sanity next time you do wiring work on one side and forgot you changed things aroud on the other side too.) Things a simple continuity tester won't show: * Breaks or bad contacts reliably if there is a fraction of connectivity left. * Swapped wires (see above). * High frequency response of the cable. Meaning it doesn't stress test and it won't show whether the cable will actually work when connected to Ethernet interfaces. But it should show the sort of trouble you seem to have. And, heres the hit, can someone in ChCh lend me one? Yes, both tester and multimeter. Volker -- Volker Kuhlmann is list0570 with the domain in header http://volker.dnsalias.net/ Please do not CC list postings to me.
Re: cable testing?
Volker Kuhlmann wrote: And, heres the hit, can someone in ChCh lend me one? Yes, both tester and multimeter. In my experience it is usually not the cables / wires which give problems but bad or incorrect terminations. I can lend you a good terminating tool if you like (not the cheap plastic one). Rob
Re: chroot sftp users
Absolutely. A backup isn't a backup unless it can be restored from. GC On Tue, Mar 2, 2010 at 5:18 PM, Jim Cheetham j...@gonzul.net wrote: On Tue, Mar 2, 2010 at 4:53 PM, Glenn Cogle gco...@gmail.com wrote: Box is backed up to tape 5 nights/week. I like backups - they help me sleep. I used to sleep because of backups ... but now I sleep because restore works. You are testing restore regularly I hope ... :-| -jim
Re: cable testing?
Nick Rout wrote, On 02/03/10 21:49: I have run a couple of cat5e cables and I am trying to terminate them, unsuccessfully at present. What kind of cable? Solid core or stranded? Is it fixed in the wall or is it running around the floor? Will a cable tester help me? I suspect that each time I put a plug on the end one or more of the wires is in the wrong place, or not quite long enough to make the connection. Coupled with this I am only 90% sure which cable end is which at the switch end (ie the centre of the star), having failed to mark them. Easy way is to use a tone source. I have one. Is there some sort of cable tester that can, eg, tell me what wires are right and what are wrong, and which end of the cable is wired wrong? Yes and no... Cable continuity testers are $20 to $100ish and they step across 8 or 9 pins sequentially, putting a voltage on the other wires to complete the circuit. More expensive boxes are called scanners and cost thousands - they can do things like testing all OSI layers (ie speak a full smtp session with a remote host) as well as testing actual voltages, cable lengths, noise factors, etc. We had an issue with an IBM x3350 server... loverly 1RU box with a dual core CPU, lightpath, raid, dual PSUs etc etc. Worth around $7k. It would randomly drop HD0, and possibly lock up and die. Once it completely lost both ethernet interfaces on a dual port PCIe NIC. It is dual UPSed up the wazoo. We replaced the IBM motherboard, drives three times, cables, backplane, raid controller, ethernet card, riser board etc all to no avail. Turns out that there was a POE injector up the cable, and it was putting out-of-spec voltages down the ethernet straight into the firewall. Putting a fuse in place (actually a cheapie 5 port ethernet switch as a sacrificial protection) seems to have cured the problem. So a voltage-reading scanner scope would have found that much quicker. The previous box in the same position was an IBM x306, which died in a similar but more permanent way. http://www.atecorp.com/equipment/fluke/675.asp That one even measures token ring round-trip time and corrupt packet generation. Its worth around $25k. And, heres the hit, can someone in ChCh lend me one? Of course! -- Craig Falconer
Re: chroot sftp users
On Wed, Mar 3, 2010 at 9:28 AM, Glenn Cogle gco...@gmail.com wrote: A backup isn't a backup unless it can be restored from. On that subject, because of O'Reilley's current 3-for-2 book deal, I just refreshed my 11 year old copy of Unix Backup Recovery with the newer Backup Recovery, which covers Unix, Linux, OS X and Windows. Very highly recommended, an excellent and accessible book. http://oreilly.com/catalog/9780596102463/ It's primarily about open source tools being used for backup environments, and is pretty close to being an essential read. However, you can get most of his advice on http://backupcentral.com/ I was flipping through the old book, and out of the front dropped some printouts :- the complete procedure for installing AMANDA (client and server) on one of my old employer's systems, and the config files! I guess in the pre-wiki world (this would have been late 1999) I'd put a copy of the procedures in the best place to look in an emergency -- in the Backup Recovery book! Now I must remember to check that I have paper dumps of restore data for other machines ... -jim