Tables-for-JIRIAF

Table 1. JRM Status Overview

Name	CPU	Memory	Walltime	NodeType	Site	AliveTime
JRM1	8	256 Gi	1:00:00	CPU	Perlmutter	0:50:00
JRM2	4	100 Gi	2:00:00	CPU	Perlmutter	0:40:00

Note: The table provides an overview of the real-time metrics and static job information for JRM. The "CPU," "Memory," and "AliveTime" columns (highlighted in green) are dynamically supplied by VK, which updates the database on the control-plane of the K8s cluster. On the other hand, the "Walltime," "NodeType," and "Site" columns (highlighted in blue) are static parameters that must be specified during the submission of a SLURM job. It is essential to ensure that the "Walltime" aligns with the duration defined in the SLURM job script.

Table 2. User-Requested Job Resources and Information

User	Name	CPU	Memory	Time	NodeType	Site	App	JobType	State
U1	Job1	5	100 Gi	0:30:00	CPU	Perlmutter	WF1	Stream	Running
U1	Job2	2	20 Gi	1:00:00	GPU	Perlmutter	WF2	Batch	Pending

Note: This table is dynamically constructed based on the specific requests made by the user, providing a comprehensive overview of the requested job resources and associated information. State information can be queried from the k8s API.

Table 2-1. File Transfer Details for Batch Type Jobs

Name	FileLocation	Method
Job2	Loc-A.tgz	scp

Note: This table presents information related to file transferring for batch-type jobs. The "Name" column identifies the job, while the "FileLocation" column specifies the location of the tarball containing all files. The "Method" column indicates the chosen transfer method, with options such as scp, globus, etc.

Table 3. Pod Status Overview

Name	CPU	Memory	Duration
Job1	5	100 Gi	0:10:00

Note: This table represents the real-time resource usage of pods, provided by VK and monitored by the Metrics Server in K8s. The "CPU," "Memory," and "Duration" columns provide insights into the current status of the respective pods