# Mystery error message from cylc/rose on macosx

**URL:** <https://cylc.discourse.group/t/mystery-error-message-from-cylc-rose-on-macosx/1329>\
**Category:** Cylc Support\
**Created:** [July 10, 2026, 11:54am UTC](https://cylc.discourse.group/t/mystery-error-message-from-cylc-rose-on-macosx/1329 "2026-07-10T11:54:19Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![gturek](https://avatars.discourse-cdn.com/v4/letter/g/a88e57/32.png) [@gturek](https://cylc.discourse.group/u/gturek)\
**Post date:** [July 10, 2026, 11:54am UTC](https://cylc.discourse.group/t/mystery-error-message-from-cylc-rose-on-macosx/1329/1 "2026-07-10T11:54:19Z")

</div>

Hi everyone. As of yesterday I keep getting this error message:

```auto
pxp-gturek:glo12_cylc gturek$ cylc install .
PluginError: Error in plugin cylc.pre_configure.rose
gaierror: [Errno 8] nodename nor servname provided, or not known: 'pxp-gturek.user.mercator-ocean.fr'

```

Same behavior with cylc-flow 3.6.4 and 3.6.5, metomi-rose 2.7.0 and 2.7.1  
but I don’t think it’s got anything to do with cylc/rose software per se, rather some (network?) setting on my machine which somehow got changed.  
Any ideas?

Thanx  
Gaby

---

<div class="post-metadata">

**Author:** ![oliver.sanders](https://yyz2.discourse-cdn.com/free1/user_avatar/cylc.discourse.group/oliver.sanders/32/110_2.png) [@oliver.sanders](https://cylc.discourse.group/u/oliver.sanders)\
**Post date:** [July 10, 2026, 3:46pm UTC](https://cylc.discourse.group/t/mystery-error-message-from-cylc-rose-on-macosx/1329/2 "2026-07-10T15:46:32Z")

</div>

Hi,

> Same behavior with cylc-flow 3.6.4

Guessing that’s 8.6.4!

This looks like a network issue. Cylc needs to be able to move around the network (via SSH) in order to submit jobs, etc. To do this, Cylc needs to be able to lookup hosts by name and expects consistency between how one host identifies itself and how another host identifies it.

I think what’s happening here is that the host identified itself as `pxp-gturek.user.mercator-ocean.fr`, but when Cylc tried to lookup this hostname, no such host was found.

Cylc uses `socket.getfqdn` to determine a host’s name:

```auto
$ python -c 'import socket; print(socket.getfqdn())'

```

And `socket.gethostbyname_ex` to lookup a host _by_ name, e.g:

```auto
$ python -c 'import socket; socket.gethostbyname_ex("pxp-gturek.user.mercator-ocean.fr")'

```

You can use those commands to test the theory. You may need to get the network people to look at it.

* * *

Aside; I think if you try running a workflow with no `rose-suite.conf` file that submits it’s jobs to another host on the network, you’ll likely hit the same error, just in a different place.

---

<div class="post-metadata">

**Author:** ![gturek](https://avatars.discourse-cdn.com/v4/letter/g/a88e57/32.png) [@gturek](https://cylc.discourse.group/u/gturek)\
**Post date:** [July 15, 2026, 8:20am UTC](https://cylc.discourse.group/t/mystery-error-message-from-cylc-rose-on-macosx/1329/3 "2026-07-15T08:20:36Z")

</div>

Thanx for that, I had my suspicions…I’ll talk to our sysadmins and report back

---

<div class="post-metadata">

**Author:** ![gturek](https://avatars.discourse-cdn.com/v4/letter/g/a88e57/32.png) [@gturek](https://cylc.discourse.group/u/gturek)\
**Post date:** [July 15, 2026, 2:45pm UTC](https://cylc.discourse.group/t/mystery-error-message-from-cylc-rose-on-macosx/1329/4 "2026-07-15T14:45:35Z")

</div>

Yep, it was decided to rm my laptop’s DNS entry. Go figure. All good now