Andyj12's questions

I have a simple IoT temp sensor (from China) and I can change the destination it routes traffic to. I want it to continue sending to their system AND to mine (redundancy, resilience etc). The sensor does not support sending to 2 destinations.

I use cloudflare for my webserver - so I have an A rule that routes traffic for a subdomain on to my IoT server. Can I use this to duplicate the traffic so it goes to BOTH the Chinese server AND my own server?

Or would I have to try setting up something that will receive data on my server and then re-send it to China (less preferable as clearly less resilient).

I have a web service (php, sql etc) running on an AWS account. I have openVPN setup, so users must establish the VPN connection, then the IP address will work.

I currently have this set as example123.noip.com which goes to 172.xxx.xxx.123

And this works.

My service is taking off, and I want to be able to offer it to other firms, BUT i want to be able to keep multiple firms hosted on the same AWS instance, but I would like to give each company a seamless experience, so for example I am thinking i can setup more noip addresses: clientcompany1.noip.com -> 172.xxx.xxx.123 clientcompany2.noip.com -> 172.xxx.xxx.123 clientcompany3.noip.com -> 172.xxx.xxx.123

So Mr X from Company3 goes to clientcompany3.noip.com and it will show his my service pages branded as Company 3... (they will not be aware other companies are hosted on that aws account)

(and some point I might need to add a separate AWS instance, and then I can just flip clientcompany3.noip.com -> 172.xxx.xxx.new address 456)

BUT - what is the best way to detect or route this within AWS? Do I just do some redirection detection from the HTTP headers using PHP? Is there a clever way to have traffic routed in with openVPN?

Sending and receiving SMS with mySMS using just bash & curl.

(Breaking the rules here - don't have a question - but I found this a challenge to make work, with few good examples out there, so here is an example. If you can improve on it - I'm certainly interested to learn something new!).

mysms.com provides a free Android only app (iPhone don't open up access to messaging), and their API allows you to send and receive SMS on that android device, receiving API calls over the web. So my use-case scenario is I want my service to be able to send out appointment reminders, and receive any replies. You need an API key - just email [email protected] and I got my API key pretty quickly.

To send SMS

1. Install the app from Play store and assign a password in the app
2. Email support and get an API key
3. code as below...

Mobile numbers need to be in full msisdn format - so country code, omit initial zero, then mobile number, so i'm running this in New Zealand which is country code +64, and mobiles are all 021, 022 or 027...

apikey="insert your api key here"
from="6421xxxxx"
pwd="you set the password in the android app"
to="6421yyyyyyy"

smsurl="https://api.mysms.com/json/message/send?api_key=$apikey&msisdn=$from&password=$pwd&recipient=$to&message=Testmessage1."

curl -v "$smsurl"

So pretty simple - think i need to do some work to cope with URL encoding of the message, but as a proof of concept this should be easy to get up and running.

Receive SMS

You need to generate an auth key to be able to access the messages and mark as read or delete. (unsure how long this is value for.)

# GET auth token  

curl -d '{ "apiKey": "'$apikey'", "msisdn": "'$from'", "password": "'$pwd'", "checkKey": "false" }' -H "Content-Type: application/json" -X POST "https://api.mysms.com/json/user/login" -o /tmp/sms.0

authtoken=$(gawk '{ i=index($0,"authToken"); x=substr($0,i+12); j=index(x,"\""); print substr(x,0,j-1); }' /tmp/sms.0)
echo "AUTH token = $authtoken"

# GET list of unread messages

curl -d '{ "apiKey": "'$apikey'", "authToken": "'$authtoken'" }' -H "Content-Type: application/json" -X POST "https://api.mysms.com/json/user/message/conversations/get" -o /tmp/sms.1


# Mark a message as read
# This one sets a message from 02123123 as having been read.
# Looks like there is a better way with MessageIDs
curl -d '{ "apiKey": "'$apikey'", "authToken": "'$authtoken'", "address": "+642123123"  }' -H "Content-Type: application/json" -X POST "https://api.mysms.com/json/user/message/conversations/read" -o /tmp/sms.2


# Deleting a message once my end has retrieved it works better for me
# so deleting a message is...
# This will delete MessageId 1
curl -d '{ "apiKey": "'$apikey'", "authToken": "'$authtoken'", "MessageId": "1"  }' -H "Content-Type: application/json" -X POST "https://api.mysms.com/json/user/message/conversations/del" -o /tmp/sms.3

I am developing a website, and I want to keep the live and some development code on the site. I was thinking of using .htaccess to allow me to develop new code, and then cut over to the new version when ready. Is this a stupid idea? and what's wrong with my .htaccess?

So, I have directories v1.1 and v1.2 in my web root. and this is my .htaccess file:

#########################

RewriteEngine On

RewriteRule ^live/(.*)$     v1.1/$1 [L]
Redirect 301 /live$ http://blah.com/live/

RewriteRule ^debug/(.*)$    v1.2/$1 [L]

# Disable directory browsing
Options All -Indexes

# Deny access to all .htaccess files
<files ~ "^.*\.([Hh][Tt][Aa])">
order allow,deny
deny from all
satisfy all
</files>

I've got this to work, and it allows example.com/live/ to show the code in v1.1 and example.com/debug to let me play with the code in v1.2.

There are 2 issues with this:

1. The user sees the word /live/ in the URL and may go digging...
2. I can't get the .htaccess file to redirect example.com/ to v1.1/

I would prefer to have example.com/<almost anything> to go to example.com/v1.1/
and example.com/debug/<anything> to go to example.com/v1.2/.

I tried...

RewriteRule ^$     v1.1/ [L]

and

RewriteRule ^(.*)$ v1.1/$1 [L]

but these did not work... (Internal Server Error).

I had a play with this excellent tool - https://htaccess.madewithlove.be/ - which seems to think my htaccess file is great (when obviously its not)

I have successfully set-up SoftEther on my AWS free tier machine, and put a small apache2 instance on there. I can access the webpage via VPN from windows, Mac & Linux.

BUT - when the VPN client is activated, ALL the traffic goes through the VPN. This is not what I intended, probably not secure, and will cost me (data throughput on AWS).

I have activated SecureNAT but I don't understand how to configure it. If I disable (basic)NAT then I can't access the web server.

The basic AWS machines has eth0: flags=4419<UP,BROADCAST,RUNNING,PROMISC,MULTICAST> mtu 1500 inet 172.31.9.151 netmask 255.255.240.0 broadcast 172.31.15.255

and the SecureNat and SoftEther setup are unchanged.

I think the issue is that the DNS gateway (as below from secureNAT) allows access not only to the webserver, but also EVERYTHING else on the web.

[SEE SecureNat standard settings][1] [1]: https://isstatic.aoverflow.com/hgcj5.png

However if I remove this, then I don't have the knowledge to add in the extra routing to JUST allow access to the webserver.

The solution has to be on the server side, not the client (as clients could alter their setup, and it will be hard to get them to use anything other than just an ovpn link anyway). However the SecureNat dialog does have client routing rules it can push (- which I haven't been able to set correctly.

Any anyone shed any light on this?