Wiki source code of Working on Maxwell
Version 13.4 by sndueste on 2025/02/05 15:10
Show last authors
author | version | line-number | content |
---|---|---|---|
1 | === Link to the [[ MAXWELL homepage >>https://docs.desy.de/maxwell/]] === | ||
2 | |||
3 | ===== ===== | ||
4 | |||
5 | |||
6 | some of the docu below may be | ||
7 | |||
8 | |||
9 | ===== Maxwell Resources ===== | ||
10 | |||
11 | The [[Maxwell Hardware >>https://docs.desy.de/maxwell/||shape="rect"]]is partitioned in different partions which will be upgraded constantly. The concept is that all integrated hardware is available via the {{code language="none"}}ALL{{/code}} partition ({{code language="none"}}ALLGPU{{/code}} for nodes with integrated GPU), but the job will be terminated when a job with a higher priority (from the nodes owner) comes in. Dedicated to Photon Science are the {{code language="none"}}PS{{/code}} (internal users) and {{code language="none"}}PSX{{/code}} (external users) partitions. Regarding GPUs (Mai 2019) the PS partition has nodes with a single NVIDEA Tesla P100 while the PSX partition has no GPUs at all. | ||
12 | |||
13 | For interactive code development (including several NVIDIA Quadro M6000 ) and running short jobs (as shared resource) the Maxwell Display Server is foreseen. Dedicated nodes will be available via [[Slurm Jobs>>url:https://confluence.desy.de/display/IS/Running+Jobs+on+Maxwell||shape="rect"]] which are not meant to work with a GUI (output to file). | ||
14 | |||
15 | Software on Maxwell is managed via the Module system. So check | ||
16 | |||
17 | |||
18 | (% class="code" %) | ||
19 | ((( | ||
20 | module avail | ||
21 | ))) | ||
22 | |||
23 | for all available software and | ||
24 | |||
25 | |||
26 | (% class="code" %) | ||
27 | ((( | ||
28 | module load * | ||
29 | ))) | ||
30 | |||
31 | e.g | ||
32 | |||
33 | |||
34 | (% class="code" %) | ||
35 | ((( | ||
36 | module load matlab/R2019a | ||
37 | ))) | ||
38 | |||
39 | to install the listed software. | ||
40 | |||
41 | ===== Access Points ===== | ||
42 | |||
43 | The most convenient and officially preferred way is via the [[Maxwell Display Server >>url:https://max-display3.desy.de:3389||shape="rect"]]via (% class="twikiNewLink" %)FastX3(%%) client or web browser. Required resources will be automatically given to new {{code language="none"}}Scientific User Accounts{{/code}} and can be requested for existing accounts via FS-EC. | ||
44 | |||
45 | [[Photon Science staff >>url:https://confluence.desy.de/display/IS/Maxwell+for+Photon+Science||shape="rect"]]can also connect to the work group server [[max-fsc.desy.de>>url:http://max-fsc.desy.de||shape="rect"]] or - if you need a GPU for your calculations - [[max-fsg.desy.de>>url:http://max-fsg.desy.de||shape="rect"]]. [[External Users>>url:https://confluence.desy.de/display/IS/Maxwell+for+Photon+Science+users||shape="rect"]] can also SSH to [[desy-ps-cpu.desy.de>>url:http://desy-ps-cpu.desy.de||shape="rect"]]{{code language="none"}}{{/code}} or - if you need a GPU for your calculations - [[desy-ps-gpu.desy.de>>url:http://desy-ps-gpu.desy.de||shape="rect"]] from inside the DESY network. From outside they have to connect to [[max-display.desy.de>>url:http://max-display.desy.de||shape="rect"]] first. | ||
46 | |||
47 | For working with [[ Python>>doc:FLASHUSER.Data Acquisition and controls.DAQ and controls overview.Online data analysis.Anaconda Python at FLASH.WebHome]]'s Jupyter Notebooks DESY provides a [[Jupyterhub>>url:https://confluence.desy.de/display/IS/JupyterHub+on+Maxwell||shape="rect"]] instance and its resource is automatically allocated to new Scientific User Accounts. | ||
48 | |||
49 | ===== During the beamtime ===== | ||
50 | |||
51 | For online processing during a beamtime it is possible to make a node reservation via maxwell(at)service, which have to be done 1 to 2 weeks before the beamtime. At the point when you start the beamtime in the [[GPFS>>url:https://confluence.desy.de/display/ASAP3/Directory+Structure||shape="rect"]] all user should have a [[DOOR>>url:https://door.desy.de/door/||shape="rect"]] account, be rgeistered as participants for the beamtime and have a scientific user account to work on this node. Currently (Mai 2019) FS could also provide a node with a single P100 GPU. | ||
52 | |||
53 | ===== Scientifc User Accounts ===== | ||
54 | |||
55 | For working at Maxwell you need the corresponding resources which the Scientific User Account. On Maxwell you can check the resources of a User via (passing the User's id is optional): | ||
56 | |||
57 | |||
58 | (% class="code" %) | ||
59 | ((( | ||
60 | my-resources id | ||
61 | ))) | ||
62 | |||
63 | When you login in Maxwell the first time a home directory with a hard quota of 20 GB is created. By default Inhouse staff will be granted the PS-resource while external users will be granted the PSX-resource. | ||
64 | |||
65 | To create a Scientific User Account the Local Contact have to request it via FS-EC while passing User's {{code language="none"}}Full Name{{/code}} and(% class="WYSIWYG_TT" %) E-mail Address(%%). FS-EC will send a form the User have to sign and after that the Local Contact gives the User (verbally) the initial password. | ||
66 | |||
67 | ===== SLURM scheduler ===== | ||
68 | |||
69 | Slurm is an open source resource management and job scheduling system for Linux clusters. It schedules via a "who comes first" basis, but is using a "back-fill algorithm". | ||
70 | |||
71 | Job submission nodes are available via the load balancer: max-display, max-fsc & max-fsg. | ||
72 | |||
73 | [[image:attach:slurm-job-submission.png||height="250"]] | ||
74 | |||
75 | |||
76 | ====== Example: Slurm Job Submission ====== | ||
77 | |||
78 | [[image:attach:slurm_job.png||height="400"]] | ||
79 | |||
80 |