Do we need Port Forwarding for STORJ V3?

Node Operators FAQ
1

I am cleaning up my port forwarding tables and discovered that I am still forwarding ports 22200 and 28967 from when I was running V2. I am running V3 (and ONLY V3) now.

Do these port forwarding rules need to be in place still for V3?

Before anyone yells at me: I did look and read many webpages on this site and others before asking here; however, I found incomplete or conflicting information.

Thanks.

2

28967 is still needed.

2 Likes
3

22200 is not needed.
Though, you still need port 28967.

1 Like
4

Just for a joke:

  1. remove one 22200 and see what happens to the node… (not needed - my node works only on 28967)
  2. then remove 28967 and see what happens to the node :wink:
    Then get the second back turned on.
1 Like
5

Sadly, I cannot do that testing because of the way they have been BANNING PEOPLE if their node isn’t up 99.97% of the time.

1 Like
6

Not true, as disqualification for downtime is disabled at the moment.

2 Likes
7

That’s good to know – because I’ve seen so many people pull out of Storj because of banning when they lost their internet or a hard drive… I’ve felt like I’m living on borrowed time, as it takes hours for my domain and IP to sync when my IP changes… I’d hate to leave the project after all this time. I joined up when it was discussed on the “Steal This Show” podcast years ago…

1 Like
8

Why is that. Are you doing this manually? Setup DDNS on your router and it will be updated automatically when your IP changes.

9

It takes time to propagate the CNAME across the internet.

10

So, you’re not using a DDNS service? Sounds like self-inflicted pain…

A DDNS record has a TTL of 60 usually. Maybe you can lower your TTL in your DNS setting if it is possible.

2 Likes
11

DDNS typically uses A or AAAA (IPv6) records in DNS and depending the DNS host, you can configure the TTL manually. Mine is 5 minutes if I remember correctly. Also, I use a script that runs every 10 minutes. It validates what IP addresses I have (v4 and v6) against what is registered in CloudFlare’s DNS (as I use CloudFlare for my DNS hosting). If it’s not up-to-date, the script sends a command to CloudFlare’s API to update the record. At worst, I’m down for maybe 15-30 minutes. Then again, my ISP tends to not change my v4 address on me, but the v6 prefix can be flaky.

12

This is incorrect. dq for downtime is not enabled.