Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Dome computer connections

  • Please log in to reply
31 replies to this topic

#1 tjensen

tjensen

    Mercury-Atlas

  • -----
  • topic starter
  • Posts: 2,712
  • Joined: 16 Feb 2005
  • Loc: Chapel Hill, NC

Posted 27 June 2020 - 08:49 AM

Hey all,

I ran an Ethernet cable out to my observatory to directly connect it to the home network since it's right on the edge of my wifi range. Well... connection speed is in the toilet. I can download 125 Gb data to a USB 3 external drive in about 20 minutes and it estimates 3 hours for the same data set via the Ethernet cable to transfer to the office computer. Are there any cable gurus out there that could help trouble shoot this?

Thanks

Tim



#2 D_talley

D_talley

    Soyuz

  • *****
  • Posts: 3,754
  • Joined: 07 Jul 2005
  • Loc: Albuquerque, New Mexico

Posted 27 June 2020 - 09:07 AM

First question is how far was the ethernet run?  I assume you have a gigabit network switch that you are connecting to. 


  • santafe retiree likes this

#3 santafe retiree

santafe retiree

    Viking 1

  • *****
  • Posts: 711
  • Joined: 23 Aug 2014

Posted 27 June 2020 - 09:26 AM

+1 to D_talley --

 

First -- ethernet distance limit is 100 meters unless you put in a powered  switch - my practice is to put a powered switch every 75 to 80 meters if the run exceeds 100 meters

 

Second - did you terminate the cables or are they pre-made?  Run a continuity test - you don't need anything fancy - this will do the job:

 

https://smile.amazon...t/dp/B072LJYHKP



#4 Couder

Couder

    Ranger 4

  • *****
  • Posts: 395
  • Joined: 24 Feb 2014
  • Loc: Missouri Ozarks

Posted 27 June 2020 - 09:33 AM

You might try

1) ethernet extenders. you plug it in an outlet closer to the observatory than the router.

2) powerline adapters - I use these in our church. Plug an ethernet cable into one in your house close to the router, plug the other one in your observatory and run cat5 cable to a switch. Muse be on the same side of the 200V on breaker.  https://www.dell.com...cB&gclsrc=aw.ds

 

3) A better router. I can pick up my signal from the observatory, and the signal is going through 2 interior and one exterior walls - 310 feet. I have another building I use wireless ethernet extender.



#5 tjensen

tjensen

    Mercury-Atlas

  • -----
  • topic starter
  • Posts: 2,712
  • Joined: 16 Feb 2005
  • Loc: Chapel Hill, NC

Posted 27 June 2020 - 09:46 AM

Thanks guys... The run is short... only 150 feet (50 m). I was connected wirelessly but data transfer rates were slow, that's why I put in the cable. The switch is a Netgear GS305 powered Gigabit. Cables were premade and rated for outside use. I'll order the tester and see what that says. Maybe I got a bad cable or it died on me. It seemed to work fine the first day I put it in. If it is the cable, that would at least explain things. Though I don't want to have to dig it up and replace it.



#6 santafe retiree

santafe retiree

    Viking 1

  • *****
  • Posts: 711
  • Joined: 23 Aug 2014

Posted 27 June 2020 - 10:06 AM

If you have direct line of sight to the obsy from your home you might want to consider a long range AP - I have good luck with these on ranches and farms:

 

https://smile.amazon...S/dp/B00HXT8FFI

 

And if you do end up digging up the cable do yourself a favor and get some 3/4" pvc pipe from your local home improvement big box, put a pull string through it and then bury it  -- that protects the cable and gives you a way to pull a new cable if needed without digging it all up

 

Cheers,

 

Tom


Edited by santafe retiree, 27 June 2020 - 10:12 AM.

  • tjensen and EFT like this

#7 Couder

Couder

    Ranger 4

  • *****
  • Posts: 395
  • Joined: 24 Feb 2014
  • Loc: Missouri Ozarks

Posted 27 June 2020 - 04:57 PM

I second what Viking 1 said - I am using an older version of the Access Point, have been using it for about 10 years. I buried my 220V line in a 2" PVC and also ran a shielded video cable in a smaller PVC. 



#8 D_talley

D_talley

    Soyuz

  • *****
  • Posts: 3,754
  • Joined: 07 Jul 2005
  • Loc: Albuquerque, New Mexico

Posted 27 June 2020 - 05:05 PM

At my old observatory I ran four strands of fiber between it and the house.  No issue with speed and it is cheap.  At my current location I have power line adapters running to the observatory and I get 550 mbps speeds to it over ethernet.  Wireless was hit or miss and speeds were very slow. 



#9 tjensen

tjensen

    Mercury-Atlas

  • -----
  • topic starter
  • Posts: 2,712
  • Joined: 16 Feb 2005
  • Loc: Chapel Hill, NC

Posted 28 June 2020 - 07:24 AM

If you have direct line of sight to the obsy from your home you might want to consider a long range AP - I have good luck with these on ranches and farms:

 

https://smile.amazon...S/dp/B00HXT8FFI

 

And if you do end up digging up the cable do yourself a favor and get some 3/4" pvc pipe from your local home improvement big box, put a pull string through it and then bury it  -- that protects the cable and gives you a way to pull a new cable if needed without digging it all up

 

Cheers,

 

Tom

Interesting idea. Might be worth having regardless... I have a CNC machine in the shop that I control with a laptop that doesn't quite register on the wireless network.



#10 tjensen

tjensen

    Mercury-Atlas

  • -----
  • topic starter
  • Posts: 2,712
  • Joined: 16 Feb 2005
  • Loc: Chapel Hill, NC

Posted 28 June 2020 - 08:05 AM

Just ran iperf3 and got a transfer speed of 940Mb/sec

 

Does that sound right?

 

Still says it will take 5 h to transfer 172 Gb


Edited by tjensen, 28 June 2020 - 08:28 AM.


#11 santafe retiree

santafe retiree

    Viking 1

  • *****
  • Posts: 711
  • Joined: 23 Aug 2014

Posted 28 June 2020 - 09:31 AM

I think your iperf test shows that the cabling is not the issue -  940mbs is near wire speed for cat 5/6 with a gigabit switch

 

So in the ideal world your transfer should take 30 minutes -- but wire speed is not the same as transfer rate -

 

Transfer rate is dependent on wire speed but also on disk speed -  which in turn is dependent on the disks themselves, at both ends, and the I/O rate of the motherboards the disks are connected to

 

You say "I can download 125 Gb data to a USB 3 external drive in about 20 minutes" -- I suggest running the USB3 transfer on each machine to see which one is the bottleneck - it may be time for an SSD or an M2 drive is you really want the fastest transfers

 

You may also want to www.fast.com on each machine for a quick and dirty speed test verification

 

You are on the right track --


  • tjensen and EFT like this

#12 tjensen

tjensen

    Mercury-Atlas

  • -----
  • topic starter
  • Posts: 2,712
  • Joined: 16 Feb 2005
  • Loc: Chapel Hill, NC

Posted 28 June 2020 - 11:03 AM

Both machines are running SSD hard drives. First upgrade I do to a new computer :)

 

Fast.com gives me internet speed, but won't tell me my home network speed will it? In any case, my internet speed was 10MBPS on both machines which is the fastest I can get in my neck of the woods

 

I'm wondering if there is a Win 10 setting that I'm missing that could be causing the bottleneck?

 

BTW, I appreciate all the help.

 

Cheers

T



#13 santafe retiree

santafe retiree

    Viking 1

  • *****
  • Posts: 711
  • Joined: 23 Aug 2014

Posted 28 June 2020 - 01:06 PM

I recommended fast.com as a quick and dirty test just to make sure there was nothing glaringly wrong with the connection to the obsy - too bad you only get the 10Mbps but at least we know the obsy end is getting at least that -- the iperf test you ran is obviously a better measure but nothing wrong with getting some verification

 

I still think you should run the USB3 data transfer rate test on both computers to make sure it is not an I/O issue at the motherboard - even though you installed SSDs on both boxes, that does not guarantee that the I/O rate on each of the MoBos is up to par

 

If that checks out then we are getting into the weeds here but the next step is to see if there is an issue at the router --

 

One way to test that is to eliminate the router entirely by giving your two computers static IP addresses on the same subnet and with the same subnet mask and the same gateway (albeit a fictitious one since there is no physical gateway under this test scenario)

 

Then connect them directly through the switch only - you will need to go through the switch in order to avoid using a cross over cable - the GS305 has MDX capability so your existing Ethernet cable will do the job - run iperf if you want but I would just do a straight data transfer to get a measure -- if the transfer speed is better then it is the router - if it is the same then I am back to wondering about I/O transfer speeds at the motherboard

 

Cheers,

 

Tom


  • tjensen likes this

#14 tjensen

tjensen

    Mercury-Atlas

  • -----
  • topic starter
  • Posts: 2,712
  • Joined: 16 Feb 2005
  • Loc: Chapel Hill, NC

Posted 28 June 2020 - 01:18 PM

Just did and actual transfer: actual transfer of 7.46 Gb took 14 mim 3 sec or 8.9 Mbps



#15 tjensen

tjensen

    Mercury-Atlas

  • -----
  • topic starter
  • Posts: 2,712
  • Joined: 16 Feb 2005
  • Loc: Chapel Hill, NC

Posted 28 June 2020 - 01:20 PM

How does one run a USB3 speed test?



#16 santafe retiree

santafe retiree

    Viking 1

  • *****
  • Posts: 711
  • Joined: 23 Aug 2014

Posted 28 June 2020 - 01:22 PM

OK, because the USB3 data transfer tests out on both platforms then the next step is to eliminate the router from the equation


  • tjensen likes this

#17 tjensen

tjensen

    Mercury-Atlas

  • -----
  • topic starter
  • Posts: 2,712
  • Joined: 16 Feb 2005
  • Loc: Chapel Hill, NC

Posted 28 June 2020 - 03:25 PM

OK. I'll give that a go tomorrow morning. I'll be busy the rest of today.



#18 santafe retiree

santafe retiree

    Viking 1

  • *****
  • Posts: 711
  • Joined: 23 Aug 2014

Posted 28 June 2020 - 03:28 PM

Cheers!  Tom



#19 NearVision

NearVision

    Vostok 1

  • -----
  • Posts: 160
  • Joined: 23 Sep 2017
  • Loc: SE Wisconsin

Posted 28 June 2020 - 07:01 PM

Have you run a traceroute? That will tell you the actual path between the 2 computers. It should go from computer 1 directly to computer 2 if your network is setup properly. The reason I ask is the speed you have in post #14.

 

Just did and actual transfer: actual transfer of 7.46 Gb took 14 mim 3 sec or 8.9 Mbps

That looks like your internet speed is a limiting factor. Another thing to look for is some older switches and routers will use the speed of the slowest thing connected to the network for everything. I haven't seen this for a while so it may not matter, but if there is an old printer or some other piece connected to the network it may be affecting things. On my home gigabit network I get around 125MB transfer rate on large files. That MB is mega-bytes not mega-bits so it works out to about the same as gigabit. 8 bits to a byte and a little overhead for error correction checking. I've also got some slower/older/smaller computers that will never get near the 125MB rate because of the internal data rates.

 

If the cable is bad check the new cable before you bury it by connecting it and trying a transfer. I agree with burying PVC and pulling the cable through it. Saves wear on the cable and makes replacing it easier.

 

Happy hunting!


  • tjensen likes this

#20 tjensen

tjensen

    Mercury-Atlas

  • -----
  • topic starter
  • Posts: 2,712
  • Joined: 16 Feb 2005
  • Loc: Chapel Hill, NC

Posted 29 June 2020 - 09:03 AM

Tried to set up the assigned IP addresses and it didn't work... no internet connection and a ping only returned 50% of the packets. If I plugged the router back into the switch, it did work but iperf returned 15 Mbits ps. So I think it switched over to wifi? Anyway, set everything back up to DHCP and ran iperf again and got 95 Mbitps. Is it possible the ethernet adapter on the PC has gone bad? Checked the cable with the Cable tester and it seems to be ok.

Was wondering if there might be a setting in the router that could be causing the issues, but I don't have any knowledge of how those things should be set so that's a dead end.

Think I might just throw in the towel on this one...



#21 tjensen

tjensen

    Mercury-Atlas

  • -----
  • topic starter
  • Posts: 2,712
  • Joined: 16 Feb 2005
  • Loc: Chapel Hill, NC

Posted 29 June 2020 - 09:06 AM

There is a trace route feature on the router, but it just gives a time so I don't know how to interpret that.

 

Edit: ran it from Win and there was 1 hop


Edited by tjensen, 29 June 2020 - 09:09 AM.


#22 NearVision

NearVision

    Vostok 1

  • -----
  • Posts: 160
  • Joined: 23 Sep 2017
  • Loc: SE Wisconsin

Posted 29 June 2020 - 10:00 AM

You would need to run the trace on 1 of the computers you are trying to transfer data between.

First: You will need the IP address of the computer that you are tracerouting to. You can find that in the network information or at a command prompt by typing 'ipconfig' without the quotes.

Next: On the other computer open a command prompt and type 'tracert "IP addr"' where "IP addr" is the IP address of the other computer. Again this is all without quotes.

 

Here are some links that may help:

https://www.whatismy...-on-windows-10/

https://support.micr...lems-in-windows


  • tjensen likes this

#23 tjensen

tjensen

    Mercury-Atlas

  • -----
  • topic starter
  • Posts: 2,712
  • Joined: 16 Feb 2005
  • Loc: Chapel Hill, NC

Posted 29 June 2020 - 10:23 AM

Yep, that's what I did.



#24 santafe retiree

santafe retiree

    Viking 1

  • *****
  • Posts: 711
  • Joined: 23 Aug 2014

Posted 29 June 2020 - 10:29 AM

Tried to set up the assigned IP addresses and it didn't work... no internet connection and a ping only returned 50% of the packets.

Because the router was removed there will be no internet -- there is no path out to the world - the only path is between the two computers -

 

Which is the point of the exercise - to isolate the two computers and to do a straight data transfer between the two in order to make sure the router is not the problem - I am concerned that you only got a 50% packet return -

 

To simplify your test,

 

1) Connect the two computers directly to each other using only the Netgear GS305 switch, your buried cable and no router

 

2) Give the two devices static addresses, e.g. Computer 1 has an address of 192.168.1.50 and Computer 2 has an address of 192.168.1.51, subnet masks of 255.255.255.0 and gateways of 192.168.1.1

 

3) From the device with address of 192.168.1.50 ping the device with the address of 192.168.1.51 - if you get 100% return with a latency of 1ms or less then the cabling/network card is probably not the issue -- if you get less than 100% return or a high latency then cabling or the network card is most likely the culprit

 

If cabling/network card is not the issue then time to look at the router - but that is another story

 

I am not persuaded by the high iperf number -- iperf sends out very small packets that is not indicative of real world data transfers of much larger packets

 

Another way to test the router is to bring the obsy computer back into the house, place it next to the router and run your wifi test from there -- then run the test with a new short ethernet cable and compare the results -

 

Don't give up -- it is just a process of elimination to narrow down the culprit  :>)

 

And flag your ping command with a -t  e.g. ping 192.168.1.51 -t   the -t flag will keep the ping going for more than the default 4 pings so you can jiggle cables, check both ends of the connection, get a 5 minute picture of packet traffic,  etc


Edited by santafe retiree, 29 June 2020 - 10:49 AM.

  • tjensen and NearVision like this

#25 tjensen

tjensen

    Mercury-Atlas

  • -----
  • topic starter
  • Posts: 2,712
  • Joined: 16 Feb 2005
  • Loc: Chapel Hill, NC

Posted 29 June 2020 - 05:17 PM


 

1) Connect the two computers directly to each other using only the Netgear GS305 switch, your buried cable and no router

 

2) Give the two devices static addresses, e.g. Computer 1 has an address of 192.168.1.50 and Computer 2 has an address of 192.168.1.51, subnet masks of 255.255.255.0 and gateways of 192.168.1.1

 

3) From the device with address of 192.168.1.50 ping the device with the address of 192.168.1.51 - if you get 100% return with a latency of 1ms or less then the cabling/network card is probably not the issue -- if you get less than 100% return or a high latency then cabling or the network card is most likely the culprit

OK, did that... one ping test returned 1 of 3 packets. When I put the -t switch in, they all failed.

 

I'll have to track down another cable. In the mean time, I have a laptop I can take down and hook up and see what happens with that.

 

Thanks for all the help and being so patient with me!!!




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics