# Nextcloud (Legacy) 33.0.8:0 fails to start after 0.4.0.1 upgrade — get-ssl-certificate: unknown hostname (valid public IP)

**URL:** <https://community.start9.com/t/nextcloud-legacy-33-0-8-0-fails-to-start-after-0-4-0-1-upgrade-get-ssl-certificate-unknown-hostname-valid-public-ip/5724>\
**Category:** Services Support\
**Created:** [September 10, 2026, 12:05am UTC](https://community.start9.com/t/nextcloud-legacy-33-0-8-0-fails-to-start-after-0-4-0-1-upgrade-get-ssl-certificate-unknown-hostname-valid-public-ip/5724 "2026-09-10T00:05:03Z")\
**Posts on this page:** 6\
**Page:** 1

<div class="post-metadata">

**Author:** ![ktthekiwi](https://community.start9.com/letter_avatar_proxy/v4/letter/k/e5b9ba/32.png) [@ktthekiwi](https://community.start9.com/u/ktthekiwi)\
**Post date:** [September 10, 2026, 12:05am UTC](https://community.start9.com/t/nextcloud-legacy-33-0-8-0-fails-to-start-after-0-4-0-1-upgrade-get-ssl-certificate-unknown-hostname-valid-public-ip/5724/1 "2026-09-10T00:05:04Z")

</div>

Hi, hoping someone can help diagnose this.

**Environment:**

- StartOS 0.4.0.1
- Nextcloud (Legacy) 33.0.8:0 — used as my primary backup destination for my computer

**Problem:**  
Nextcloud shows “Running” but the Web Interface health check never clears — it’s been stuck on “Nextcloud is initializing” for several days. Opening the UI gives ERR\_CONNECTION\_CLOSED in the browser.

**Log excerpt from the service:**

```auto
Error in host RPC: Error: {"method":"get-ssl-certificate", ... "error":{"code":6,"message":"Not Found","data":{"details":"unknown hostname: 125.236.192.37", ...
Location: shared-libs/crates/start-core/src/service/effects/net/ssl.rs:50

```

**What I’ve checked:**

- 125.236.192.37 is my genuine current public IP address (confirmed via an external IP lookup) — it is not stale or outdated.
- The Health Checks panel separately shows a script error on every check:  
`/usr/local/bin/check-web.sh: line 3: /dev/stdin: No such device or address`  
`/usr/local/bin/check-web.sh: line 4: ((: <= 5000: syntax error: operand expected (error token is "<= 5000")`
- I’m not blocked on a version-skip issue — under Marketplace \> Nextcloud \> Version, the available versions are 34.0.3:6, 33.0.8:3, 33.0.7:0, and 32.0.11:0. I’m already on 33.0.8:0, past the oldest listed version, so there’s nothing missing in the sequence.

**Why I’m posting rather than experimenting:**  
This looks like it matches other threads reporting the same error signature — get-ssl-certificate rejecting a correctly-registered public IP, on the same “(Legacy)” package and StartOS version. I have a large amount of important data in this Nextcloud instance, so I don’t want to uninstall, reinstall, or rebuild the package without guidance.

**Questions:**

1. Is this a known issue with the SSL certificate system in 0.4.0.1?
2. Would updating to the available 33.0.8:3 build (same major version, later build) be safe to try, or is it likely to help with this specific error?
3. Is there a safe way to resolve this without risking my existing data?

Thanks in advance.

---

<div class="post-metadata">

**Author:** ![StuPleb](https://community.start9.com/user_avatar/community.start9.com/stupleb/32/195_2.png) [@StuPleb](https://community.start9.com/u/StuPleb)\
**Post date:** [September 10, 2026, 1:28pm UTC](https://community.start9.com/t/nextcloud-legacy-33-0-8-0-fails-to-start-after-0-4-0-1-upgrade-get-ssl-certificate-unknown-hostname-valid-public-ip/5724/2 "2026-09-10T13:28:39Z")

</div>

> [@ktthekiwi](#):
>
> - Nextcloud (Legacy) 33.0.8:0

I didn’t read past this line. Why are you trying to start a Legacy service on StartOS v040? Anyone starting a Legacy service has not followed the update guide and is _asking_ for problems, including possible data loss.

---

<div class="post-metadata">

**Author:** ![cgal](https://community.start9.com/letter_avatar_proxy/v4/letter/c/ebca7d/32.png) [@cgal](https://community.start9.com/u/cgal)\
**Post date:** [September 11, 2026, 6:09pm UTC](https://community.start9.com/t/nextcloud-legacy-33-0-8-0-fails-to-start-after-0-4-0-1-upgrade-get-ssl-certificate-unknown-hostname-valid-public-ip/5724/3 "2026-09-11T18:09:12Z")

</div>

Hey guys. I’m having the same issue. I’m on NextCloud 31.0.13 and just updated to Start9 0.4.0. I’m unable to start the service due to this issue and I’m also unable to update to NextCloud 34.0.3. Here’s a more comprehensive write up, with some obfuscation  
Any troubleshooting help would be greatly appreciated. I did take a backup of the services before updating, but if it’s something that can be pretty easily fixed with a CLI command or something I’d much prefer that. If I’m able to update the service and _then_ it solves the ssl issue, I’m all ears. Just want to be back in a working state.

Thank you!

**Title:** Nextcloud cannot start on StartOS 0.4.0: get-ssl-certificate fails with unknown hostname = gateway WAN IPv4

**Environment**

- StartOS 0.4.0 (upgraded; first-run / early 0.4 use)
- Nextcloud package **31.0.13:1** (Marketplace also offers **33.0.8:3** )
- Server on LAN \<LAN\_IPv4\>/24, default gateway \<LAN\_GATEWAY\_IPv4\>
- Router: ASUS RT (StartOS reports no UPnP/PCP/NAT-PMP)
- mDNS: embassy-.local
- Detected WAN IPv4: **\<WAN\_IPv4\>** (wan\_ip: Some(…) on eno1 / Wired connection 1)
- Not CGNAT (public IPv4 is visible to StartOS)

**Summary**  
Nextcloud never reaches a healthy start. On start, the package calls get-ssl-certificate with every address it knows, including the gateway WAN IPv4. StartOS returns Not Found: unknown hostname: \<WAN\_IPv4\> (ssl.rs:50) and start aborts.

The WAN address appears on **Nextcloud → Interfaces → Wired connection 1** as Public IPv4 (https://\<WAN\_IPv4\>:). That row cannot be deleted. Enabling or disabling it does not register the name with the cert signer.

**Steps**

1. Install / keep Nextcloud 31.0.13:1 on 0.4.0.
2. Interfaces show Local IPv4 + Local mDNS enabled; Public IPv4 present.
3. Start Nextcloud.

**Expected**  
Start succeeds using LAN / mDNS / onion (and only _enabled_ public names). Disabled or unregistered WAN IPv4 should not be required on the cert.

**Actual**

text

```auto
Error in host RPC: ... method: get-ssl-certificate
hostnames: [nextcloud.embassy, 10.0.3.1, 127.0.0.1, ::1, fd00:3::1,
  <link-local-IPv6-bridge>, <onion>, <LAN_IPv4>,
  embassy-<id>.local, <link-local-IPv6-nic>, <WAN_IPv4>]
error: unknown hostname: <WAN_IPv4>
Location: shared-libs/crates/start-core/src/service/effects/net/ssl.rs:50

```

**Tried (no change)**

- Public IPv4 on Nextcloud Interfaces: off and on
- System IPv4 / WAN enabled
- start-cli net ssl generate-certificate \<WAN\_IPv4\> \<LAN\_IPv4\> embassy-.local nextcloud.embassy (PEM printed; same RPC error on next start)
- UI session error after toggle: RPC ERROR: Invalid Signature replay attack detected (stale session after net change; not an actual attack)

**Second issue (do not “fix” by jumping majors)**  
Marketplace update **31.0.13:1 → 33.0.8:3** fails as designed:

text

```auto
Error: cannot migrate from 31.0.13:1 to 33.0.8:3

```

Nextcloud never completed a healthy 31 start, so 32/33 should not be applied until this SSL start failure is fixed. Please publish or point to a **32.x** bridge after 31 runs.

**Gateway sync excerpt (system logs)**  
eno1 / Wired connection 1: subnets: {\<LAN\_IPv4\>/24, /64}, lan\_ip: {\<LAN\_GATEWAY\_IPv4\>}, wan\_ip: Some(\<WAN\_IPv4\>), gateway\_type: InboundOutbound. PCP / NAT-PMP / UPnP: supported: false.

---

<div class="post-metadata">

**Author:** ![StuPleb](https://community.start9.com/user_avatar/community.start9.com/stupleb/32/195_2.png) [@StuPleb](https://community.start9.com/u/StuPleb)\
**Post date:** [September 11, 2026, 6:51pm UTC](https://community.start9.com/t/nextcloud-legacy-33-0-8-0-fails-to-start-after-0-4-0-1-upgrade-get-ssl-certificate-unknown-hostname-valid-public-ip/5724/4 "2026-09-11T18:51:56Z")

</div>

You CANNOT jump across various versions. If you’re very far being on your updates such that you’re on v31 and the latest is v34, you must install v32 first, use it, then install v33, use it, and only then install v34. The error message when you try to jump versions explains this.

How you could be on `v31 (Legacy)` and having just updated to StartOS v040 I don’t understand though. The latest version of Nextcloud on StartOS v0351 is v33.0.8. At one of more stages you haven’t followed the update guide. You’re supposed to update your services before you update StartOS.

The rest of the post, I’m not sure we should be looking at, since it sounds like you’re trying to run v31-legacy on StartOS v040 – there’ll be all kinds of errors.

---

<div class="post-metadata">

**Author:** ![cgal](https://community.start9.com/letter_avatar_proxy/v4/letter/c/ebca7d/32.png) [@cgal](https://community.start9.com/u/cgal)\
**Post date:** [September 12, 2026, 4:38pm UTC](https://community.start9.com/t/nextcloud-legacy-33-0-8-0-fails-to-start-after-0-4-0-1-upgrade-get-ssl-certificate-unknown-hostname-valid-public-ip/5724/5 "2026-09-12T16:38:45Z")

</div>

Hey Stu, for whatever reason I thought I had updated all my apps on 3.5.x. Upon restoring my backup I can confirm that that assumption was incorrect. I’ll run through the upgrade a little more gracefully this time and I bet it’ll work just fine. Most importantly, my backup restored beautifully, so there was no data loss. Thanks for your time

---

<div class="post-metadata">

**Author:** ![ktthekiwi](https://community.start9.com/letter_avatar_proxy/v4/letter/k/e5b9ba/32.png) [@ktthekiwi](https://community.start9.com/u/ktthekiwi)\
**Post date:** [September 16, 2026, 9:21am UTC](https://community.start9.com/t/nextcloud-legacy-33-0-8-0-fails-to-start-after-0-4-0-1-upgrade-get-ssl-certificate-unknown-hostname-valid-public-ip/5724/6 "2026-09-16T09:21:23Z")

</div>

I uninstalled Nextcloud and then reinstalled. All working beautifully again. No data was lost.
