HTCondor - Revision history

Cacosta: /* What happens if my ticket expires? */

2026-10-21T04:17:16Z

What happens if my ticket expires?

Cacosta: /* Kerberos authentication on User Interfaces */

2026-10-19T04:10:42Z

Kerberos authentication on User Interfaces

Cacosta: /* Kerberos authentication on User Interfaces */

2026-10-19T04:09:32Z

Kerberos authentication on User Interfaces

Cacosta: /* Kerberos authentication in User Interfaces */

2026-10-19T04:08:31Z

Kerberos authentication in User Interfaces

Cacosta: /* Kerberos authentication on User Interfaces */

2026-10-19T04:07:13Z

Kerberos authentication on User Interfaces

Cacosta: /* Kerberos authentication in User Interfaces */

2026-10-19T04:06:59Z

Kerberos authentication in User Interfaces

Cacosta: /* Kerberos authentication in User Interfaces */

2026-10-19T04:06:25Z

Kerberos authentication in User Interfaces

Cacosta: /* Submitting your jobs */

2026-10-19T04:04:14Z

Submitting your jobs

Vacin: /* Interactive submission */

2026-08-06T12:49:34Z

Interactive submission

Vacin: /* Requests */

2026-08-06T12:38:39Z

Requests

@@ Line 150: / Line 150: @@
 -- Failed to fetch ads from: <XXXXXX> : machine.pic.es
 AUTHENTICATE:1003:Failed to authenticate with any method
 AUTHENTICATE:1004:Failed to authenticate using KERBEROS
 AUTHENTICATE:1004:Failed to authenticate using PASSWORD

@@ Line 110: / Line 110: @@
 = Kerberos authentication on User Interfaces =
-In order to interact with HTCondor (e.g., submitting jobs or querying the queue) on the User Interfaces, you need a valid Kerberos ticket.
+In order to interact with HTCondor (e.g., submitting jobs or querying the queue) on the User Interfaces, you need a valid Kerberos ticket. Note that this only affects UIs. A Kerberos ticket is not required to submit jobs from within a Jupyter Notebook session.
 Each time you SSH into a ui.pic.es machine (or a specific ui from your experiment), a new Kerberos ticket is automatically generated for you. Therefore, in practice, you shouldn't notice any real change in your workflow.
@@ Line 140: / Line 140: @@
 	renew until 10/21/26 06:43:20
 </pre>
 == What happens if my ticket expires? ==

@@ Line 140: / Line 140: @@
 	renew until 10/21/26 06:43:20
 </pre>
 == What happens if my ticket expires? ==

@@ Line 108: / Line 108: @@
 </pre>
-= Kerberos authentication in User Interfaces =
+= Kerberos authentication on User Interfaces =
-In order to interact with HTCondor (e.g., submitting jobs or querying the queue) in the User Interfaces, you need a valid Kerberos ticket.
+In order to interact with HTCondor (e.g., submitting jobs or querying the queue) on the User Interfaces, you need a valid Kerberos ticket.
 Each time you SSH into a ui.pic.es machine (or a specific ui from your experiment), a new Kerberos ticket is automatically generated for you. Therefore, in practice, you shouldn't notice any real change in your workflow.

@@ Line 108: / Line 108: @@
 </pre>
-= Kerberos authentication on User Interfaces =
+= Kerberos authentication in User Interfaces =
 In order to interact with HTCondor (e.g., submitting jobs or querying the queue) in the User Interfaces, you need a valid Kerberos ticket.

@@ Line 112: / Line 112: @@
 In order to interact with HTCondor (e.g., submitting jobs or querying the queue) in the User Interfaces, you need a valid Kerberos ticket.
-Each time you SSH into a ui.pic.es machine, a new Kerberos ticket is automatically generated for you. Therefore, in practice, you shouldn't notice any real change in your workflow.
+Each time you SSH into a ui.pic.es machine (or a specific ui from your experiment), a new Kerberos ticket is automatically generated for you. Therefore, in practice, you shouldn't notice any real change in your workflow.
-Important Note regarding running jobs: The Kerberos ticket is ONLY required to authenticate your commands (like condor_submit or condor_q). Once your job is successfully submitted to the queue, it will run normally until completion. Your jobs will NOT fail or be interrupted if your Kerberos ticket expires while they are running.
+Important Note regarding running jobs: The Kerberos ticket is ONLY required to authenticate your commands (like condor_submit or condor_q). Once your job is successfully submitted to the queue, it will run normally until completion. Your jobs will not fail or be interrupted if your Kerberos ticket expires while they are running.
 You can check the validity of your current ticket with the klist command:
@@ Line 128: / Line 128: @@
 </pre>
-The standard ticket expires in 24 hours and it is renewable for up to 5 days. "Renewable" means that you can extend the life of your Kerberos ticket without needing to introduce your credentials again, simply by running:
+The standard ticket expires in 24 hours and it is renewable for up to 5 days. "Renewable" means that you can extend the life of your Kerberos ticket without needing to enter your credentials again, simply by running:
 <pre>
@@ Line 156: / Line 156: @@
 </pre>
-To solve this, you just need to create a new ticket running the kinit_pic command or just log in again by SSH to the machine. You will be prompted to authenticate via web login.
+To solve this, you just need to create a new ticket by running the kinit_pic command or just log in again via SSH to the machine. You will be prompted to authenticate via web login.
 For instance, running kinit_pic:
@@ Line 171: / Line 171: @@
 	renew until 10/23/26 06:49:20
 </pre>
 = Submitting your jobs =

@@ Line 107: / Line 107: @@
 All jobs in cluster 281 have been marked for removal
 </pre>
 = Submitting your jobs =

@@ Line 750: / Line 750: @@
 == Interactive submission ==
-As it happens with Torque/Maui, HTCondor also has the possibility to submit interactive jobs using the option "-i" or "-interactive".
+HTCondor has the possibility to submit interactive jobs using the option "-i" or "-interactive".
 <pre>

← Older revision		Revision as of 12:38, 6 August 2026
Line 353:		Line 353:


−	You can access any of these machines directly by using the request_gpus option. However, please note that jobs running on gpu06-gpu12 will be in preemptible mode. This means that if a higher-priority job (from Jupyter or one of the priority groups) requires the resources, your job may be terminated and re-queued. The ~~execptions~~ are <b>gpu05, gpu13-gpu15</b>, where submitted jobs will not be preempted by any other job once they are running.	+	You can access any of these machines directly by using the request_gpus option. However, please note that jobs running on gpu06-gpu12 will be in preemptible mode. This means that if a higher-priority job (from Jupyter or one of the priority groups) requires the resources, your job may be terminated and re-queued. The exceptions are <b>gpu05, gpu13-gpu15</b>, where submitted jobs will not be preempted by any other job once they are running.

	Example of submit file to run in any possible gpu:		Example of submit file to run in any possible gpu: