Error when user tries to log in.

182 views
Skip to first unread message

Robert Henning

unread,
Jun 8, 2022, 5:04:28 PM6/8/22
to Discuss
I've been using globus for many years with very few problems but this error popped up today when someone tried to access their data. Can anyone tell me how to fix it. Unfortunately, I'm still running V4. 


Command Failed: Error (login)
Endpoint: /mnt/beegfs/data/chen_2206/ (2fe8f6f8-e696-11ec-9bd1-2d2219dcc1fa)
Server: helix.cars.aps.anl.gov:2811
Message: Login Failed
---
Details: 530-Login incorrect. : globus_credential: Error verifying credential: Failed to verify credential
530-globus_gsi_callback_module: Could not verify credential
530-globus_gsi_callback_module: The certificate is not yet valid: Cert with subject: /C=US/O=Globus Consortium/OU=Globus Connect Service/CN=cb1d8162-9a10-11e6-b0c2-22000b92c261/CN=useridb is not yet valid- check clock skew between hosts.
530 End.      


Thanks,

Rob

Jack Kordas

unread,
Jun 8, 2022, 5:14:42 PM6/8/22
to Robert Henning, Discuss
Hi Rob,

Have the endpoint admin make sure that ntp or some other time synchronization service is running  on the endpoint.  It appears that its clock is behind.  If you wait a while after activation, the error may go away as well.

 - Jack

Robert Henning

unread,
Jun 9, 2022, 9:39:58 AM6/9/22
to Discuss, kor...@globus.org, Discuss, Robert Henning
Hi Jack,

Thanks for the response. I checked the ntp settings on the server and found that it was unsynchronized. I restarted the service and it was synchronized again but it still didn't work. I let it go overnight but it is still not working. 

If you or anyone else has any ideas, I would appreciate it. 

Thanks,

Rob

Michael Link

unread,
Jun 9, 2022, 11:32:39 AM6/9/22
to Discuss, Robert Henning
Hi Rob,

I looked into your endpoint and it looks like this is an issue we've
seen before on some endpoints that started out on an early GCSv4 version
and then renewed the original MyProxy CA certificate.

The easiest fix is to remove or #comment out the
`certificate_issuer_subca_certfile` line from
/var/lib/globus-connect-server/myproxy-server.conf, and then restart the
myproxy-server service.
You'll need to recreate the share.

If you still have issues after that, please follow up at sup...@globus.org.

Mike
> <http://helix.cars.aps.anl.gov:2811>

Michael Link

unread,
Jun 9, 2022, 3:52:52 PM6/9/22
to dis...@globus.org
Great, happy to help.

Mike

On 6/9/2022 1:28 PM, Robert Henning wrote:
> This worked!
>
> Thank you very much!
>
> Rob
>
> On Thu, Jun 9, 2022 at 10:44 AM Michael Link <ml...@globus.org
> <mailto:ml...@globus.org>> wrote:
>
> One other thing before recreating the share -- make sure you deactivate
> the host endpoint from the details page and are forced to log in
> again to
> MyProxy server.  Otherwise the cert from the old configuration may
> get used.
>
> Mike
>
> On 6/9/2022 10:32 AM, Michael Link wrote:
> > Hi Rob,
> >
> > I looked into your endpoint and it looks like this is an issue we've
> > seen before on some endpoints that started out on an early GCSv4
> version
> > and then renewed the original MyProxy CA certificate.
> >
> > The easiest fix is to remove or #comment out the
> > `certificate_issuer_subca_certfile` line from
> > /var/lib/globus-connect-server/myproxy-server.conf, and then
> restart the
> > myproxy-server service.
> > You'll need to recreate the share.
> >
> > If you still have issues after that, please follow up at
> > sup...@globus.org <mailto:sup...@globus.org>.
> >
> > Mike
> >
> >
> >
> > On 6/9/2022 8:39 AM, Robert Henning wrote:
> >> Hi Jack,
> >>
> >> Thanks for the response. I checked the ntp settings on the
> server and
> >> found that it was unsynchronized. I restarted the service and it
> was
> >> synchronized again but it still didn't work. I let it go
> overnight but
> >> it is still not working.
> >>
> >> If you or anyone else has any ideas, I would appreciate it.
> >>
> >> Thanks,
> >>
> >> Rob
> >>
> >> On Wednesday, June 8, 2022 at 4:14:42 PM UTC-5 kor...@globus.org
> <mailto:kor...@globus.org> wrote:
> >>
> >>     Hi Rob,
> >>
> >>     Have the endpoint admin make sure that ntp or some other time
> >>     synchronization service is running  on the endpoint.  It appears
> >>     that its clock is behind.  If you wait a while after
> activation, the
> >>     error may go away as well.
> >>
> >>       - Jack
> >>
> >>     On Wed, Jun 8, 2022 at 4:04 PM Robert Henning
> <henn...@gmail.com <mailto:henn...@gmail.com>>
> >> wrote:
> >>
> >>         I've been using globus for many years with very few
> problems but
> >>         this error popped up today when someone tried to access
> their
> >>         data. Can anyone tell me how to fix it. Unfortunately,
> I'm still
> >>         running V4.
> >>
> >>
> >>         Command Failed: Error (login)
> >>         Endpoint: /mnt/beegfs/data/chen_2206/
> >>         (2fe8f6f8-e696-11ec-9bd1-2d2219dcc1fa)
> >>         Server: helix.cars.aps.anl.gov:2811
> <http://helix.cars.aps.anl.gov:2811>
> >>         <http://helix.cars.aps.anl.gov:2811
Reply all
Reply to author
Forward
0 new messages