Yesterday I learned about postgrey and what it does to help protect you from spam. Basically it delays any email that it doesn’t recognize by causing a soft bounce. Many spam senders won’t send a second time. After it sees the sending server a few times it will white list the server.
With Amazon SES that isn’t so easy. Amazon uses many different servers to send emails on behalf of many businesses. It was taking sometimes over 5 minutes for emails to reach me when doing testing. I found something on a forum that helped me white list all of amazon’s servers. Got my delay time down to a second or two.
If you are running your own mail server I am sure you are comfortable poking around in it. If you ssh in and go to /etc/postgrey/ you will find two files. One is whitelist_clients and another is whitelist_receipients. You want to edit the whitelist_clients.
At the very end you need to add: /^.*.amazonses.com/
I reloaded postgrey and then tested and it worked like a charm.
This has bugged me for far too long. I think at one point I did have this resolved but then I hosed my pbx and when I restored it I forgot what I had done. In any event, the problem for me was converting the Freepbx information to Wazo.
When you read the write-up by Ward Mundy it mentions that security is handled by registering your IP with telecomsxchange first. That only verifies the IP addresses that can access their system. Also, in the Freepbx list it doesn’t mention anything about usernames and passwords.
I kept getting a ‘Failed to authenticate on INVITE’ message in my logs. Lots of time wasted but now I have it resolved. Your Telecomsxchange profile includes a username and password. You do need to enter that in the SIP Trunk section of Wazo. From there the settings pretty much line up.
In the outgoing calls section is where you will use the Extern prefix box for your external prefix. I did put my “1” after the # in extern prefix as that is how it is listed on the TelecomsXchange site.
I want to explain a little more on how to use Signalwire with Asterisk and Wazo. I had written a post a few months back on how to make calls with Signalwire but I never got the receive part working correctly. I think this time I have cracked the code.
These instructions should be the same for Twilio.
You will be making two entries for the trunk. One is for the inbound and one is for the outbound.
The settings are each the same except that you will be registering one of the trunks and the other trunk you will not be.
For registration put in the name of your trunk in both name and authentication username. Password is the trunk password. Remote server is the xxx.sip.signalwire.com. Port is 5060. For the contact field, and this is very important, you need to enter in a verified phone number on your account. For example +15125551212
Signaling and advanced are the same settings I had listed before.
Remember, one of the trunks you fill out the register field and the other trunk you don’t register.
Lastly, for inbound calls be sure to put a 1 in front of the phone number otherwise you will get the message about an invalid destination.
On the signalwire side just make sure you choose custom ciphers and disable the first one. AEAD_AES_256_GCM_8
I wasn’t sure how to title this. Basically, I just want to clarify some inconsistencies in their documentation. When they say “Base URL” they have it listed as https://example.signalwire.com/api/laml/2010-04-01. Well, I couldn’t get that to work for me when I was trying to use it with Textit. I had to use https://example.signalwire.com/api/laml/2010-04-01/Accounts
Also, your AuthToken is your API token. Something you generate. I figured that out easily enough and it is a better implementation than Twilio‘s secret but the API base URL I really thought should have been listed with “Accounts” at the end of it.
Is free openheatmap.com better than Excel? For me it was by far. I used Excel to clean up my data but then it’s heat map options were just awful. I could barely tell there was a numerical difference between each state. I tried other colors but to no avail.
I took the data and imported it into openheatmap.com and bam! Worked like a charm. I was trying to plot some data on the 50 US States and was just having a miserable time doing it Excel. Excel isn’t free. Excel isn’t cheap. Sure, I could have tried libreoffice – which I did – but had some issues with all the data. 10,000+ rows is too much for a pivot table in a 32 bit application. Had to uninstall and install a 64 bit version.
For me, openheatmap was awesome. Putting your data into a visual representation has always been more helpful than displaying columns and rows of data. In this instance I was actually surprised at the results that it showed. The data will be used to direct marketing money to different states.
After all these years VPN technology is still frustrating. I spent hours yesterday and today trying to setup two different VPNs. One was IPSEC site to site and the other was openvpn.
The Site to Site was started yesterday. I was between to Sophos XG routers. I wasted quite a bit of time setting it up only to find out that both devices needed to be on the same firmware. This morning I updated the firmware on both routers and I was able to connect. Stability seemed fine on my side but the connection on the other end seemed a bit sluggish.
This leads me to OpenVPN. Have used OpenVPN for several years but only off and on. I have setup a number of OpenVPN servers but always on linux.
The goal for this setup was to setup a High Availability (HA) Cloud-based Asterisk system with a local Asterisk system. Apparently the key to get the master to fail over to the slave is they both need to be on the same subnet. I got them to sync but I couldn’t force the failover.
Working with this garbage called Windows Server 2008 I have failed miserably. I don’t have a linux server to use on this site. Well, I do but it’s already running the PBX in the VM and that is what has become highly unstable. Not sure why as I have the same setup working great in other locations. Hyper-V on this server is also very poor.
I also want to say that I am not even going to bother with PPTP. It’s never been reliable enough.
My passenger window would not roll down. I already knew why I just couldn’t remember what I needed to do to fix it.
Just like my IT and entertainment troubleshooting I want to leave this here so I am remember next time.
When the battery is disconnected on virtually any car it wipes the radio, clock and messes with the automatic windows.
The passenger window will work with the passenger side controls but not on the drivers side. Simply roll the passenger window down (using passenger side controls) just a little bit and then roll it back up by holding the button in the up position for a few seconds. This will reset the system and your window will work again.
I believe my wife’s Ford Fusion has the same commands to fix her windows.
Conversion failed. The transcoder crashed or failed to start up was annoying message for me. Thankfully it looks like it was an easy fix for me. Apparently all the movies I had on my Plex server I must have removed. I added one back and then it worked. However, before I did that I checked the permissions of the transcode folder and found no issues.
I was unable to find my plex server for a very long time. I got frustrated and left it. It took me SIX months to get back to it. SIX MONTHS! I was frustrated. Thankfully I have it figured out now.
My issue was one of simplicity and stupidity but the latter might be too harsh. My issues were actually two fold. Number 1, I had 2 devices that were registering with the same IP. Not sure how that happened but I had one set as a static DHCP and another as just a static on the device. Switched them both to static DHCP with the Plex server having a new IP. Rebooted and that was that. The second issue I came across – and this one took awhile to discover – was that I had my server under one plex account email and my plex clients under another account. I’m using ubuntu server to host so I have to edit an XML file. It varies depending on what OS you are using.
What’s weird for me is the name of the support article. It’s called “Why am I locked out of server settings and how do I get back in?” I wouldn’t have thought to use any of those keywords in a search. I get how it is the same issue but still weird for me.
I had been unable to call landlines for some strange reason. It was in the US Virgin Islands. My trunk provider is voip.ms. Kept getting a busy signal. I troubleshooted it with voip.ms and as always they found a way.
Very simple put, my caller ID was set to area code + number. 3407741340. This caller ID worked great for any US numbers that I tried and even my 340 cellphone. I had another system setup where I was also able to call landlines. This had me believe the problem was in my dialplans. Nope, I was wrong again. If I moved the trunk out of my asterisk system and used it as a stand alone SIP trunk I had the same issue.
The solution was very simple and something I hadn’t considered. I was missing a “1” in front of the phone number listed in the caller ID. That’s it. That is all it was..