[World Community Grid] hitparade van week 40

[KuuKe]

Inhoudsopgave World Community Grid hitparades week 40

• DPC World Community Grid hitparade van 28 september 2025 - geen stats
• DPC World Community Grid hitparade van 29 september 2025
• DPC World Community Grid hitparade van 30 september 2025 - geen stats
• DPC World Community Grid hitparade van 1 oktober 2025 - geen stats
• DPC World Community Grid hitparade van 2 oktober 2025
• DPC World Community Grid hitparade van 3 oktober 2025
• DPC World Community Grid hitparade van 4 oktober 2025 - geen stats

TEAM gegenereerd met: Weektopic TEAM Generator

[ Voor 5% gewijzigd door KuuKe op 04-10-2025 17:04 ]

[KuuKe]

Update!!!

September 30, 2025
Tech team is working through issues with the WAS container configuration and the BOINC daemon builds on the science cluster.
We are also pinging data centre admis as we don't have the /science filesystem on 1/6 of the science nodes, don't have the ability to add the IP address to the NFS share
Hopefully, not a blocker but does lock that node out of some roles until we can mount the NFS share.

[KuuKe]

En weer een update:

October 2, 2025
Hosting resolved the issue with our NFS shares in Openstack; scheduler and download path seem to work; we are debugging the new containerized file_upload_handler FCGI worker pool and file_upload_handler integration with Kafka through the librdkafka library.
We will remove the ACL restrictions that are causing 403 forbidden when users try to contact the scheduler or upload files shortly.
We are still validating the end-to-end upload path with our local BOINC clients. Once we confirm it, we will open up traffic to the upload servers, start serving the downloads that were available when we shutdown, and we can start turning the BOINC daemons back on.

[DerManiak]

Zou 't dan toch echt weer gaan lopen? Inmiddels ruim een maand verder, voor een migratie waarvoor ze zelf 48 uur hadden ingeschat
Op dit moment krijg ik nog steeds de 403-error, benieuwd of vandaag de uploads gaan lukken.

[KuuKe]

October 3, 2025
We are aware of the issue with the scheduler returning "Another scheduler instance is running for this host" and have identified the cause in the config.xml template we adapated for the new containerzied environment. We will fix it once we have confirmed that the new event-driven validation and assmilation pipelines are working correctly.
Uploads are being processed normally, we've confirmed the new architecture for the containerized file_upload_handler pool behind Apache is correctly producing to the per-application Kafka (Redpanda) topics, storing the event and result data in separate queues on the local brokers partition.
As a result, there will be at least one more weekend sprint. Tentatively, we expect to be producing new workunits next week for MCM1, ARP1, and MAM1 beta version 7.07, validations should resume over the weekend, initial releases of batches will be intermittent.