Hi all,
Continuing my slack conversation earlier today (for those who have access: https://hpctoolset.slack.com/archives/C016SDKRUE8/p1770055124953129), I think I found an issue with xdmod container when a newer version becomes available out there. I got the hpc-toolset-tutorial docker compose environment pulled & luanched pretty fresh (few days ago). In .env file: HPCTS_VERSION="2025.02" .
Upon examining the log of docker today, I found this:
(2026-01-30 4:33:43): [be[default]] [sbus_issue_request_done] (0x0040): sssd.dataprovider.getAccountInfo: Error [1432158212]: SSSD is offline
(2026-01-30 4:33:43): [nss] [cache_req_common_process_dp_reply] (0x3f7c0): [CID#3] CR #2: Could not get account info [1432158212]: SSSD is offline
ERROR 2002 (HY000): Can't connect to MySQL server on 'mysql' (115)
-- Waiting for database to become active ...
ERROR 2002 (HY000): Can't connect to MySQL server on 'mysql' (115)
-- Waiting for database to become active ...
---> Open XDMoD Setup: SSO...
---> Open XDMoD Setup: start
spawn xdmod-setup
You are currently using Open XDMoD 11.0.0, but a newer version
(11.0.2) is available.
Do you want to continue (yes, no)? [no] 1
'1' is not a valid option.
Do you want to continue (yes, no)? [no]
Failed to get prompt
---> Open XDMoD Setup: hpc resource
spawn xdmod-setup
You are currently using Open XDMoD 11.0.0, but a newer version
(11.0.2) is available.
Do you want to continue (yes, no)? [no] 4
'4' is not a valid option.
Do you want to continue (yes, no)? [no] 1
The responses to xdmod-setup script were all wrong because it did not answer the first one ("Do you want to continue"). I will see if there is a way to disable "new version checking" -- I saw that in the xdmod code. (Or if there is a way to modify the Expect code to anticipate that and answer "yes" if a new version appears.)
The TL;DR Origin
I am trying to stand up the HPC toolset tutorial on a VM environment in a cloud somewhere (Google cloud). The VM is not seen from the internet at large, I believe. I ssh-ed into this VM using gcloud compute ssh. I finally managed to establish the port forwarding and able to contact the ssh login to the cluster (ssh port 6222), Open OnDemand (https 3443) and ColdFront (https 2443). But I am having error contacting the XDMoD port 4443 with this error when accessing with Firefox:
Secure Connection Failed
An error occurred during a connection to localhost:4443:4443. PR_END_OF_FILE_ERROR
Error code: PR_END_OF_FILE_ERROR
The page you are trying to view cannot be shown because the authenticity of the received data could not be verified.
Please contact the website owners to inform them of this problem.
The reported error was varying in nature (for example Chrome said: "localhost unexpectedly closed the connection").
I finally found out that there was an error contacting the server process itself, and saw on the backend (inside the container) that the xdmod was never launched correctly.
Hi all,
Continuing my slack conversation earlier today (for those who have access: https://hpctoolset.slack.com/archives/C016SDKRUE8/p1770055124953129), I think I found an issue with xdmod container when a newer version becomes available out there. I got the hpc-toolset-tutorial docker compose environment pulled & luanched pretty fresh (few days ago). In
.envfile:HPCTS_VERSION="2025.02".Upon examining the log of docker today, I found this:
The responses to
xdmod-setupscript were all wrong because it did not answer the first one ("Do you want to continue"). I will see if there is a way to disable "new version checking" -- I saw that in the xdmod code. (Or if there is a way to modify the Expect code to anticipate that and answer "yes" if a new version appears.)The TL;DR Origin
I am trying to stand up the HPC toolset tutorial on a VM environment in a cloud somewhere (Google cloud). The VM is not seen from the internet at large, I believe. I ssh-ed into this VM using
gcloud compute ssh. I finally managed to establish the port forwarding and able to contact the ssh login to the cluster (ssh port 6222), Open OnDemand (https 3443) and ColdFront (https 2443). But I am having error contacting the XDMoD port 4443 with this error when accessing with Firefox:The reported error was varying in nature (for example Chrome said: "localhost unexpectedly closed the connection").
I finally found out that there was an error contacting the server process itself, and saw on the backend (inside the container) that the xdmod was never launched correctly.