Always check your new server IP addresses | Hacker News Reader

Always check your new server IP addresses

varaneckas.com

36 pointsspajus6 comments

https://mxtoolbox.com/ is a useful tool for doing this check; simply type blacklist:your_ip and you can check against all major spam blacklists.

From what I've heard and seen, DO, EC2, and Linode address blocks are more likely than not to be blacklisted by default. For transactional mail, you're much better off using something like Mailgun, Socketlabs, or Sendgrid. Some amount of marketing emails can also be sent through those services, but if you don't fit within the patterns they support, and if you're confident in taking control of your own mail reputation, you'll want to look at lesser-known VPS hosting companies whose IP blocks aren't blacklisted.

Thanks, that should really have been in the main post. But does that check also include the filters that Google and Facebook likely use? I guess Facebook doesn't use Spam mail filters?

If you qualify and are willing to pay, you can use Return Path which integrates with the big providers and has visibility into where you're placed in the inbox. But in general it can be hard to even get reports when someone clicks the Spam button in Gmail! If you're a large volume sender you might be able to get into https://support.google.com/mail/answer/6254652?hl=en but in general, for smaller senders and B2B sales outreaches, it's unlikely they'll give access. Other providers have things like https://help.yahoo.com/kb/SLN3438.html which can give you granular feedback. And of course set up and monitor your DMARC feedback addresses!

How is it acceptable to sell service on a spammed IP addresses?

Aren't providers supposed to prevent this?

Sorry/Uzuojauta.

Well, you have to remember that this isn't like DO is selling a stolen car or something, this is a bunch of third parties all deciding to reduce the value of that IP address. There's nothing inherently "wrong" with the IP address, just lots of shitty customer support.

How is it acceptable to sell service on a spammed IP addresses?

Tying assessments of trust to IP addresses through blacklisting is crude mechanism, and becomes more and more susceptible to type 1 errors as demand on the v4 address space reduces the time between address reassignment.

But it's an easy mechanism to implement, so basically nobody using it wants to admit it's flawed or learn to deal with it's unreliability.

Reply on news.ycombinator.com