Jump to content
Search In
  • More options...
Find results that contain...
Find results in...


  • Content Count

  • Joined

  • Last visited


This user doesn't have any awards

About pomtom44

  • Title

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Hi all I have a small project im doing on my off time at work, where we have aprox 50 HP thin clients (Yes i have asked HP about this, but they dont see it as a issue so wont create a solution for me for it) What im after is a way to load a background image from a webshare on boot, so we can generate a new background daily and apply it to all the units, without having to push config changes to all the units (Doesn't work if the unit is offline, etc) So far I have found the command which loads the background image, and if i run it from a terminal on the unit, it works fine wget -P /tmp/ http://int-web-server/background.jpg hptc-zero-desktop --root --image=/tmp/background.jpg - Fetch image and save to tmp - Load image into background However when I put it as a dhcp hook call, (so i know theres internet before trying to wget a file) I can see the file download, and the set background command runs, but I get a error XOpenDisplay returned NULL! I have a feeling its because the dhcp hook isnt running as the "user" so there's no background to attach the image to? but i have no idea how to get it to run as the user?
  2. I think i found it, for some reason my network card was configured to use a /32 rather than /24 No idea why that would stop only 1 port from working, but i have changed that and it appears to be working as expected now
  3. There is no firewall on either linux box (that im aware of), and no firewall between servers I have HAproxy between my WAN and servers as im learning how it works and how to loadbalance various services, I have other things which run though it (plex for example) I did have apache running as the proxy but wanted to experiment with a dedicated proxy server and try something new Ignore the fact about HAproxy as im not touching that in any way for this test (its installed on the server im testing from, but i dont go though it for curling)
  4. Im trying to setup the reverse proxy I have a HAproxy server which im trying to use to convert a hostname into host:port But HAproxy was showing the server as offline even though i could access it fine from my windows pc via the web interface so i did a curl from the HAproxy server to see what was going on and found that I can curl to my apache install on my webserver but not to my git install on the same webserver (two services, same server) Gitea is responding fine to responses from localhost and from my windows pc Just not to the HAproxy server
  5. Apache is working fine, its gitea which isnt responding (not sure what web backend it uses) And as you say, it only errors when coming from my other linux server, it works fine from windows Either way hears the output (assumed you mean run on the web/git server) Active Internet connections (only servers) Proto Recv-Q Send-Q Local Address Foreign Address State PID/Program name tcp 0 0* LISTEN 1060/sshd tcp 0 0* LISTEN 1134/lighttpd tcp 0 0* LISTEN 1237/tor tcp 0 0* LISTEN 1673/pmcd tcp 0 0* LISTEN 1674/pmproxy tcp 0 0* LISTEN 42911/pmlogger tcp 0 0* LISTEN 1076/dnsmasq tcp6 0 0 :::22 :::* LISTEN 1060/sshd tcp6 0 0 :::3000 :::* LISTEN 8684/gitea tcp6 0 0 :::443 :::* LISTEN 6322/apache2 tcp6 0 0 :::44321 :::* LISTEN 1673/pmcd tcp6 0 0 :::44322 :::* LISTEN 1674/pmproxy tcp6 0 0 :::9090 :::* LISTEN 1/init tcp6 0 0 :::4330 :::* LISTEN 42911/pmlogger tcp6 0 0 :::3306 :::* LISTEN 1052/mysqld tcp6 0 0 :::80 :::* LISTEN 6322/apache2 udp 0 0* 1076/dnsmasq udp 0 0* 1076/dnsmasq
  6. At this point im not using a proxy, thats what im trying to troubleshoot by using curl my webserver is an apache server, and git is gitea Both hosted on the same physical server There is no firewall between the proxy server (where im curling from) and the web/git server (where im curling to) Both are on the same network (as I can ping and curl to apache fine, just not git) So yes, something is misconfigured somewhere, i just have no idea where My thought would be its my proxyserver but I dont know why curl to 80 works fine but 3000 doesnt If easier, ignore the fact its a proxy and see it as Linux Server ---> Web Server Curl:80 ----> 200OK Curl:3000 -----> Timeout Curl Localhost:3000 OK
  7. administrator@server:~$ curl -v localhost:3000 * Rebuilt URL to: localhost:3000/ * Trying ::1... * Connected to localhost (::1) port 3000 (#0) > GET / HTTP/1.1 > Host: localhost:3000 > User-Agent: curl/7.47.0 > Accept: */* > < HTTP/1.1 302 Found < Content-Type: text/html; charset=utf-8 < Location: /explore < Set-Cookie: lang=en-US; Path=/; Max-Age=2147483647 < Set-Cookie: i_like_gitea=ec85572f80d33bbe; Path=/; HttpOnly < Set-Cookie: _csrf=Np3J33SmZkolOPfsidXXNRovLX46MTU4MDQyMTU0MjUwNzYyOTkwNw%3D%3D; Path=/; Expires=Fri, 31 Jan 2020 21:59:02 GMT; HttpOnly < X-Frame-Options: SAMEORIGIN < Date: Thu, 30 Jan 2020 21:59:02 GMT < Content-Length: 31 < <a href="/explore">Found</a>. * Connection #0 to host localhost left intact
  8. Windows to WebServer C:\Users\Thomas>curl -v * Rebuilt URL to: * Trying * TCP_NODELAY set * Connected to ( port 80 (#0) > GET / HTTP/1.1 > Host: > User-Agent: curl/7.55.1 > Accept: */* > < HTTP/1.1 200 OK < Date: Thu, 30 Jan 2020 21:51:47 GMT < Server: Apache < Last-Modified: Mon, 11 Nov 2019 22:43:15 GMT < ETag: "2-59719db6ec6c0" < Accept-Ranges: bytes < Content-Length: 2 < Content-Type: text/html < hi* Connection #0 to host left intact Linux to WebServer root@D-LoadBalancer:/home/administrator# curl -v * Rebuilt URL to: * Trying * Connected to ( port 80 (#0) > GET / HTTP/1.1 > Host: > User-Agent: curl/7.47.0 > Accept: */* > < HTTP/1.1 200 OK < Date: Thu, 30 Jan 2020 21:51:49 GMT < Server: Apache < Last-Modified: Mon, 11 Nov 2019 22:43:15 GMT < ETag: "2-59719db6ec6c0" < Accept-Ranges: bytes < Content-Length: 2 < Content-Type: text/html < * Connection #0 to host left intact Windows to Git C:\Users\Thomas>curl -v * Rebuilt URL to: * Trying * TCP_NODELAY set * Connected to ( port 3000 (#0) > GET / HTTP/1.1 > Host: > User-Agent: curl/7.55.1 > Accept: */* > < HTTP/1.1 302 Found < Content-Type: text/html; charset=utf-8 < Location: /explore < Set-Cookie: lang=en-US; Path=/; Max-Age=2147483647 < Set-Cookie: i_like_gitea=11b50e8c70058e2f; Path=/; HttpOnly < Set-Cookie: _csrf=3ul8XzBg8VpzWzRlhmia2QUBptg6MTU4MDQyMTE2NTAxMjUyNjc1NA%3D%3D; Path=/; Expires=Fri, 31 Jan 2020 21:52:45 GMT; HttpOnly < X-Frame-Options: SAMEORIGIN < Date: Thu, 30 Jan 2020 21:52:45 GMT < Content-Length: 31 < <a href="/explore">Found</a>. * Connection #0 to host left intact Linux to Git hiroot@D-LoadBalancer:/home/administrator# curl -v * Rebuilt URL to: * Trying * connect to port 3000 failed: Connection timed out * Failed to connect to port 3000: Connection timed out * Closing connection 0 curl: (7) Failed to connect to port 3000: Connection timed out
  9. Hi all Hopefully someone can help as im lost and not fully understanding whats going on I have a home-lab that i use to try new things out, alot of what I do is overkill for what I need, but it lets me learn and develop (thats what a homelab is for after all) I have a web server with a few websites i host, i installed gitea on it on a separate port, and run it all though haproxy to allow for port remapping based on hostname (give my git server a url that isnt a custom port) and so I can learn load-balancing later on for web and databases It was working fine, until i went to get lets encrypt working via haproxy (no idea if thats what broke it, but thats what I changed most recently ) My haproxy is now saying the git server is unreachable, and cant pass traffic though to it I tried accessing it via its IP and port and it loads fine, so i know the server is responding fine I then read on how haproxy does health checks and found it just pings it and gets a 200OK or other codes back and uses that for its status I did a curl between servers, and got 200OK back for my webserver, but when i put my port in for the git it just sits there waiting I then did a curl from my windows pc to the webserver and got 200 back from the webserver and 302 for the git server my question is, why can my proxy server not get a reply on the git port? and why was it working fine before but stopped working randomly? Can try explain more if needed.
  10. https://www.youtube.com/watch?v=NnYTCVuATZM
  11. Thanks for the input Both strips are on separate Cat5 cabling. At the moment I have Orange for Pos, Brown for Neg and Blue for data, same on each strip / cat5 My Strip is a 12V WS2811 I know it isnt a power issue to the LED's as it works fine when I run it from the Nano Unless your talking about the data signal being 5v or 12v? Its a 5v data signal, which is why I have the 3.3 to 5v logic level shifter, as the arduino outputs 3.3v and the led strip expects 5 If I run the NodeMCU directly to the first LED strip then I get glitching on the leds, but if I put the shifter in then the signal is stable and works 100%
  12. Sorry if this is something simple, but im scratching my head over this for a while and there are a few things that have me super confused. I have two WS2811 strips lighting up my outdoor stairs Both are powered from a central point, using Cat5 Cabling to get power and data from the controller to the start of each strip Iv been running it from a Nano perfectly fine for a while, but wanted to add networking so upgraded to a NodeMCU I first started on just one strip to confirm it worked as expected, and after setting about 50 of the 80 leds it started to glitch, so I added a 3.3 to 5v converter and the issues went away. I then went to work on the second strip, and was getting nothing from it, no commands sent to the strip were working, the first strip kept working fine however. I swapped around the pin outputs and converters in case it was a faulty part or bad pin, but no matter the combination, it worked 100% when using the first strip, and never worked on the second. So i put my Nano back in, running the same color swap test code I had, and both strips worked fine, put the Node back in and only the one strip worked. I then got a small test strip and connected it directly to the output of the converters and when it was on the one for the first strip, it worked fine (With both the test and first strip displaying the same test) But when I put it on the output for the second strip, the test strip would turn off as if it was getting a "Set all to 0" command. I disconnected the cat5 that connected the converter to the second strip, and the test strip started working. So I thought there must be a short or something on the cat5 So I disconnected the LED strip from the other end and did a continuity test, and none of the leads were shorting (Pos, Neg and Data) So I put the cable back onto the converter, with nothing on the other end, and the test strip stopped working again, I dont know if the length of the cat5 is causing a capacitance issue or something, even though its only connected on one end (Converter end) And it works fine with the nano, only when its the Node and only with the second strip do I have problems Hopefully my description makes sense, and someone has some idea of whats going on Link to diagram of my setup https://imgur.com/a/pTefyR5 And happy to try make a short clip if a visual guide would help explain my problem Thanks in advance
  13. Hi all Not sure if this is the best place for this question, but it seems like the only place it fits We have a problem at work that we are scratching our heads over and dr google doesn't seem to be helping at all. We are migrating from a deskphone to a softphone solution, and as such we have upgraded our system to a RDS2019 farm from our old 2008 R2 cluster (We use RDS and thinclients for 99% of our staff) We are having issues with audio quality and we are trying to emilitate parts of the chain that could be causing issues At the moment we are ignoring the thinclients entirely and testing our RDS gateway (we have two gateways for remote authentication for staff, Terminal services cluster gateway 2008 r2 and RDS Gateway farm 2019) For testing we have one of our staff offsite who remote into their workstation on site with a softphone client installed on the PC When calling locally (They are onsite) the audio quality is 100% so we can rule out any SIP / network issues When offsite using the old RDS 2008 Gateway to their workstation, the quality is about 95% (taking into account network latancy and such) When offsite using the new RDS2019 Gateway the quality drops to about 70% (They are still clearly audible, but there is definite distortion compared to the 2008 gateway) The only thing that is different on the 3 tests is the change of gateway, it all goes though the same network to the same workstation and software at the end. So we are 95% sure its something different on how the two gateways handle the audio Because we are using 2019 as I said before, google isnt the most helpful as everything comes back with 2008 or 2016, and even then its not the issues we are having. Does anyone have any ideas on what we should do to get the same quality on both gateways? Happy to explain anything in more detail or answer any other questions Thanks in advance
  14. HI all Firstly I know this is a very specific set of requirements, so im not expecting there to be much out there in the way of what i looking for. If I can't find anything that works then ill be using HAproxy as that seems to meet most of my requirements. Im looking for a http / https load balancer / proxy that I can put on my home network Currently im using apache for my web server, and I have a few proxys running on it for other services like plex. however I want to put a proper proxy in place before my web server, so I can install some failover web servers (Partly a learning experience, partly because I want to) My ideal service would be as follows - Easy to maintain / configure (I often spin up a test site for projects im working on, so not having to spend 20 mins re-configuring a proxy would be ideal) - Following on from above, a web based GUI (While I dont mind getting into config files, i find a simple gui is alot better for how I work, and i dont have to re-learn a config file structure if I dont touch it in a year) - Installed on ubuntu (as that is what my other servers run so want to keep it uniform) - Prefer not a docker image (Never was a fan of docker, im still on the old "Spin up a VM for each service" mentality) - Doesnt run anything on port 9090 (Cockpit management runs on that port) - Lets Encrypt as part of the service (Dont want to have to be running a second config / script for cert renewals, just another step id have to remember when adding more sites) As i said before if I can't find anything then HAproxy with a custom LE script would be my best bet, but doesn't hurt to ask to see whats out there Thanks in advance for any suggestions