Hi!
I am playing around and testing pangolin intensively and struggled to add more than 3 newt sites to my main instance version 1.15.0?
I looked here in the issues and found nothing, as well as in the docs and didnt find any defined limits?
I was using CE and switched over to EE and a private licence but still: limited to 3 sites (3 newt and one local. If i try to add another one then only local is selectable).
Bug or user error?
Environment
OS Type & Version: Ubuntu 22.04
Pangolin Version: dockerized 1.15 (ee-latest)
Gerbil Version: latest (1.3.0?)
Traefik Version: latest (3.6.7?)
Newt Version: 1.9.0
To Reproduce
try to add a 4th newt site
Expected Behavior
able to add more
Originally created by @valki2 on GitHub (Jan 28, 2026).
Original GitHub issue: https://github.com/fosrl/pangolin/issues/2365
### Describe the Bug
Hi!
I am playing around and testing pangolin intensively and struggled to add more than 3 newt sites to my main instance version 1.15.0?
I looked here in the issues and found nothing, as well as in the docs and didnt find any defined limits?
I was using CE and switched over to EE and a private licence but still: limited to 3 sites (3 newt and one local. If i try to add another one then only local is selectable).
Bug or user error?
### Environment
- OS Type & Version: Ubuntu 22.04
- Pangolin Version: dockerized 1.15 (ee-latest)
- Gerbil Version: latest (1.3.0?)
- Traefik Version: latest (3.6.7?)
- Newt Version: 1.9.0
### To Reproduce
try to add a 4th newt site
### Expected Behavior
able to add more
@markusbenkert commented on GitHub (Jan 28, 2026):
I believe this is stated in the license terms for private use of the EE version, that only three sites may be added.
<!-- gh-comment-id:3814409748 -->
@markusbenkert commented on GitHub (Jan 28, 2026):
I believe this is stated in the license terms for private use of the EE version, that only three sites may be added.
@oschwartz10612 commented on GitHub (Jan 28, 2026):
Hum... there should not be any limit to the number of sites or users you
can add in either edition. The only thing that changes is if you have
supported the project with a limited supporter key.
What happens when you try to add the 4th site? Could you send a screenshot?
<!-- gh-comment-id:3814514303 -->
@oschwartz10612 commented on GitHub (Jan 28, 2026):
Hum... there should not be any limit to the number of sites or users you
can add in either edition. The only thing that changes is if you have
supported the project with a limited supporter key.
What happens when you try to add the 4th site? Could you send a screenshot?
"Only local selectable" reminds me to this issue .
Could you check your DB?
<!-- gh-comment-id:3816528713 -->
@Lokowitz commented on GitHub (Jan 29, 2026):
"Only local selectable" reminds me to this [issue ](https://github.com/fosrl/pangolin/issues/1724).
Could you check your DB?
my org (there is no other):
subnet 100.90.128.0/24
util. subnet: 100.96.128.0/24
Sites subnets and adresses:
local: 0.0.0.0/32
site 1/newt: 10.0.0.64/26 100.90.128.0/24
site 2/newt: 10.0.0.128/26 100.90.128.1/24
site 3/newt: 10.0.0.192/26 100.90.128.6/24
config.yml
there is no org block, and the only cidr is in gerbil:
subnet_group: "10.0.0.0/8"
Looks like this is not the same problem like in issue 1724 as the local site is 0.0.0.0/32 and the adresses of the sites (100.90.128.X/24) match the subnet of the org (100.90.128.0/24)
<!-- gh-comment-id:3818066756 -->
@valki2 commented on GitHub (Jan 29, 2026):
@Lokowitz
Hi! Thanks for the hint - i checked my database:
Exit node:
10.0.0.1/24
my org (there is no other):
subnet 100.90.128.0/24
util. subnet: 100.96.128.0/24
Sites subnets and adresses:
local: 0.0.0.0/32
site 1/newt: 10.0.0.64/26 100.90.128.0/24
site 2/newt: 10.0.0.128/26 100.90.128.1/24
site 3/newt: 10.0.0.192/26 100.90.128.6/24
config.yml
there is no org block, and the only cidr is in gerbil:
subnet_group: "10.0.0.0/8"
Looks like this is not the same problem like in issue 1724 as the local site is 0.0.0.0/32 and the adresses of the sites (100.90.128.X/24) match the subnet of the org (100.90.128.0/24)
@oschwartz10612 Screenshot looks the same like https://github.com/fosrl/pangolin/issues/1724#issuecomment-3430519032
@oschwartz10612 commented on GitHub (Jan 29, 2026):
When you try to add the site could you look at the server logs. I think it will print out the error. I suspect you might be running out of IPs somehow.
<!-- gh-comment-id:3819791159 -->
@oschwartz10612 commented on GitHub (Jan 29, 2026):
When you try to add the site could you look at the server logs. I think it will print out the error. I suspect you might be running out of IPs somehow.
<!-- gh-comment-id:3824028964 -->
@valki2 commented on GitHub (Jan 30, 2026):
Unfortunately, there is no error (log level is debug)...
And to "try to add a site" ends already on the url https://pangolin.DOMAIN.com/SITE/settings/sites/create as there is only local available to select :)
{"data":null,"success":false,"error":true,"message":"No available subnets","status":500,"stack":null}
But looking above there are plenty of subnets left? :)
<!-- gh-comment-id:3842454036 -->
@valki2 commented on GitHub (Feb 3, 2026):
Got a good hint from a friend! Dev Tools unveil the following response of https://pangolin.DOMAIN.com/api/v1/org/ORG/pick-site-defaults:
{"data":null,"success":false,"error":true,"message":"No available subnets","status":500,"stack":null}
But looking above there are plenty of subnets left? :)
@github-actions[bot] commented on GitHub (Feb 18, 2026):
This issue has been automatically marked as stale due to 14 days of inactivity. It will be closed in 14 days if no further activity occurs.
<!-- gh-comment-id:3917819119 -->
@github-actions[bot] commented on GitHub (Feb 18, 2026):
This issue has been automatically marked as stale due to 14 days of inactivity. It will be closed in 14 days if no further activity occurs.
I will try to backup the database and write an export script for the most important stuff that is not reconfigured within minutes:
resources.
licence
Then deploy a clean pangolin container, configure everthing and reimport that stuff.
I'll keep you informed.
<!-- gh-comment-id:3998645503 -->
@valki2 commented on GitHub (Mar 4, 2026):
I will try to backup the database and write an export script for the most important stuff that is not reconfigured within minutes:
- resources.
- licence
Then deploy a clean pangolin container, configure everthing and reimport that stuff.
I'll keep you informed.
<!-- gh-comment-id:4024129318 -->
@nlsrchtr commented on GitHub (Mar 9, 2026):
Hi @valki2,
since I might be [hit by the same situation](https://github.com/fosrl/pangolin/issues/1724#issuecomment-3964810414), I was wondering if you found a solution for you?
Regards,
When resetting the database, 2 things are annoying:
public resources are a lot to type in again via GUI
your licence will be gone as the instance name changes
Nevertheless, I tried it out and succeeded so far. there are the steps if someone wants to go on the same journey:
Stop your instance
Backup your database
in my docker enviroment: killed all files in the volume except the needed files like config.yml, traefik config, plugins like middlewaremanager etc.
fire up the new, clean instance (using the install token from the log)
recreate your organization - keep the same UID!
recreate your sites - again: keep the same UIDs as well as the numeric IDs by creating them in correct order
fire up your preferred weapon of choice for sqlite (sqlite3, DB Browser for sqlite...) and open the new DB, then attach the old one:
ATTACH 'dbalt.sqlite' AS dbalt;
import the resources
PRAGMA foreign_keys = OFF;
INSERT INTO resources SELECT * FROM dbalt.resources;
INSERT INTO targets SELECT * FROM dbalt.targets;
INSERT INTO roleResources SELECT * FROM dbalt.roleResources;
INSERT INTO targetHealthCheck SELECT * FROM dbalt.targetHealthCheck;
copy that db back and fire up your new instance
Note:
Of course you have to reconnect your sites, recreate your private ressources etc. - but at least all the public resources AKA "traefik stuff" AKA reverse proxy stuff is running so the downtime of all visible resources is minimal and you got less pressure.
Note 2:
I was surprised that the new db was around 500kb while the old one is around 2.5GB! the table requestauditlog is very,very hungry.
<!-- gh-comment-id:4038303228 -->
@valki2 commented on GitHub (Mar 11, 2026):
Not a solution but a workaround:
When resetting the database, 2 things are annoying:
- public resources are a lot to type in again via GUI
- your licence will be gone as the instance name changes
Nevertheless, I tried it out and succeeded so far. there are the steps if someone wants to go on the same journey:
1. Stop your instance
2. Backup your database
3. in my docker enviroment: killed all files in the volume except the needed files like config.yml, traefik config, plugins like middlewaremanager etc.
4. fire up the new, clean instance (using the install token from the log)
5. recreate your organization - keep the same UID!
6. recreate your sites - again: keep the same UIDs as well as the numeric IDs by creating them in correct order
7. fire up your preferred weapon of choice for sqlite (sqlite3, DB Browser for sqlite...) and open the new DB, then attach the old one:
ATTACH 'dbalt.sqlite' AS dbalt;
8. import the resources
PRAGMA foreign_keys = OFF;
INSERT INTO resources SELECT * FROM dbalt.resources;
INSERT INTO targets SELECT * FROM dbalt.targets;
INSERT INTO roleResources SELECT * FROM dbalt.roleResources;
INSERT INTO targetHealthCheck SELECT * FROM dbalt.targetHealthCheck;
9. copy that db back and fire up your new instance
Note:
Of course you have to reconnect your sites, recreate your private ressources etc. - but at least all the public resources AKA "traefik stuff" AKA reverse proxy stuff is running so the downtime of all visible resources is minimal and you got less pressure.
Note 2:
I was surprised that the new db was around 500kb while the old one is around 2.5GB! the table requestauditlog is very,very hungry.
Note 3: Really write down your newt sites IDs and secrets, as it seems now you cannot regenerate those without an enterprise licence. oO
<!-- gh-comment-id:4038340867 -->
@valki2 commented on GitHub (Mar 11, 2026):
Note 3: Really write down your newt sites IDs and secrets, as it seems now you cannot regenerate those without an enterprise licence. oO
@github-actions[bot] commented on GitHub (Mar 26, 2026):
This issue has been automatically marked as stale due to 14 days of inactivity. It will be closed in 14 days if no further activity occurs.
<!-- gh-comment-id:4130754049 -->
@github-actions[bot] commented on GitHub (Mar 26, 2026):
This issue has been automatically marked as stale due to 14 days of inactivity. It will be closed in 14 days if no further activity occurs.
@LevitateDilationEggbeater commented on GitHub (Apr 3, 2026):
I am facing the issue to and would like to connect more sites without all the manual intervention.
<!-- gh-comment-id:4182387252 -->
@LevitateDilationEggbeater commented on GitHub (Apr 3, 2026):
I am facing the issue to and would like to connect more sites without all the manual intervention.
<!-- gh-comment-id:4187139766 -->
@oschwartz10612 commented on GitHub (Apr 4, 2026):
Unfortunately you will need to manually open up the subnet right now if you are out of ips: https://github.com/fosrl/pangolin/issues/1724
@LaurenceJJones commented on GitHub (Apr 14, 2026):
Since this topic is widely covered as a limitation of older pangolins (with context that we used to use a smaller range)and/or user just has lots of sites so uses the availability so in short manual intervention is needed check the linked issue above.
Marking as completed.
<!-- gh-comment-id:4245570167 -->
@LaurenceJJones commented on GitHub (Apr 14, 2026):
Since this topic is widely covered as a limitation of older pangolins (with context that we used to use a smaller range)and/or user just has lots of sites so uses the availability so in short manual intervention is needed check the linked issue above.
Marking as completed.
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Originally created by @valki2 on GitHub (Jan 28, 2026).
Original GitHub issue: https://github.com/fosrl/pangolin/issues/2365
Describe the Bug
Hi!
I am playing around and testing pangolin intensively and struggled to add more than 3 newt sites to my main instance version 1.15.0?
I looked here in the issues and found nothing, as well as in the docs and didnt find any defined limits?
I was using CE and switched over to EE and a private licence but still: limited to 3 sites (3 newt and one local. If i try to add another one then only local is selectable).
Bug or user error?
Environment
To Reproduce
try to add a 4th newt site
Expected Behavior
able to add more
@markusbenkert commented on GitHub (Jan 28, 2026):
I believe this is stated in the license terms for private use of the EE version, that only three sites may be added.
@oschwartz10612 commented on GitHub (Jan 28, 2026):
Hum... there should not be any limit to the number of sites or users you
can add in either edition. The only thing that changes is if you have
supported the project with a limited supporter key.
What happens when you try to add the 4th site? Could you send a screenshot?
@Lokowitz commented on GitHub (Jan 29, 2026):
"Only local selectable" reminds me to this issue .
Could you check your DB?
@valki2 commented on GitHub (Jan 29, 2026):
@Lokowitz
Hi! Thanks for the hint - i checked my database:
Exit node:
10.0.0.1/24
my org (there is no other):
subnet 100.90.128.0/24
util. subnet: 100.96.128.0/24
Sites subnets and adresses:
local: 0.0.0.0/32
site 1/newt: 10.0.0.64/26 100.90.128.0/24
site 2/newt: 10.0.0.128/26 100.90.128.1/24
site 3/newt: 10.0.0.192/26 100.90.128.6/24
config.yml
there is no org block, and the only cidr is in gerbil:
subnet_group: "10.0.0.0/8"
Looks like this is not the same problem like in issue 1724 as the local site is 0.0.0.0/32 and the adresses of the sites (100.90.128.X/24) match the subnet of the org (100.90.128.0/24)
@oschwartz10612 Screenshot looks the same like https://github.com/fosrl/pangolin/issues/1724#issuecomment-3430519032
@oschwartz10612 commented on GitHub (Jan 29, 2026):
When you try to add the site could you look at the server logs. I think it will print out the error. I suspect you might be running out of IPs somehow.
@valki2 commented on GitHub (Jan 30, 2026):
Unfortunately, there is no error (log level is debug)...
And to "try to add a site" ends already on the url https://pangolin.DOMAIN.com/SITE/settings/sites/create as there is only local available to select :)
@valki2 commented on GitHub (Jan 30, 2026):
Btw: adding a second org and create a site there shows the same problem.
@valki2 commented on GitHub (Feb 3, 2026):
Got a good hint from a friend! Dev Tools unveil the following response of https://pangolin.DOMAIN.com/api/v1/org/ORG/pick-site-defaults:
{"data":null,"success":false,"error":true,"message":"No available subnets","status":500,"stack":null}
But looking above there are plenty of subnets left? :)
@github-actions[bot] commented on GitHub (Feb 18, 2026):
This issue has been automatically marked as stale due to 14 days of inactivity. It will be closed in 14 days if no further activity occurs.
@valki2 commented on GitHub (Mar 4, 2026):
I will try to backup the database and write an export script for the most important stuff that is not reconfigured within minutes:
Then deploy a clean pangolin container, configure everthing and reimport that stuff.
I'll keep you informed.
@nlsrchtr commented on GitHub (Mar 9, 2026):
Hi @valki2,
since I might be hit by the same situation, I was wondering if you found a solution for you?
Regards,
@valki2 commented on GitHub (Mar 11, 2026):
Not a solution but a workaround:
When resetting the database, 2 things are annoying:
Nevertheless, I tried it out and succeeded so far. there are the steps if someone wants to go on the same journey:
ATTACH 'dbalt.sqlite' AS dbalt;
PRAGMA foreign_keys = OFF;
INSERT INTO resources SELECT * FROM dbalt.resources;
INSERT INTO targets SELECT * FROM dbalt.targets;
INSERT INTO roleResources SELECT * FROM dbalt.roleResources;
INSERT INTO targetHealthCheck SELECT * FROM dbalt.targetHealthCheck;
Note:
Of course you have to reconnect your sites, recreate your private ressources etc. - but at least all the public resources AKA "traefik stuff" AKA reverse proxy stuff is running so the downtime of all visible resources is minimal and you got less pressure.
Note 2:
I was surprised that the new db was around 500kb while the old one is around 2.5GB! the table requestauditlog is very,very hungry.
@valki2 commented on GitHub (Mar 11, 2026):
Note 3: Really write down your newt sites IDs and secrets, as it seems now you cannot regenerate those without an enterprise licence. oO
@github-actions[bot] commented on GitHub (Mar 26, 2026):
This issue has been automatically marked as stale due to 14 days of inactivity. It will be closed in 14 days if no further activity occurs.
@LevitateDilationEggbeater commented on GitHub (Apr 3, 2026):
I am facing the issue to and would like to connect more sites without all the manual intervention.
@oschwartz10612 commented on GitHub (Apr 4, 2026):
Unfortunately you will need to manually open up the subnet right now if you are out of ips: https://github.com/fosrl/pangolin/issues/1724
@LaurenceJJones commented on GitHub (Apr 14, 2026):
Since this topic is widely covered as a limitation of older pangolins (with context that we used to use a smaller range)and/or user just has lots of sites so uses the availability so in short manual intervention is needed check the linked issue above.
Marking as completed.