The four Proxmox nodes have one (SAMSUNG MZ7LH480HAHQ 480G) SSD each with disk wear at 79, 63, 73, 72%. They are mostly consumed by (local storage) kube VMs.
Do we have any cold spares ready?
How does the replacement work? Wait for 99% wearout? Offline dd? Half-online zfs mirror?
Kube local VMs seem to not be able to migrate to other nodes.
Kindly stop requiring local storage for kube applications :)
This should be automatically tracked.
The four Proxmox nodes have one (SAMSUNG MZ7LH480HAHQ 480G) SSD each with disk wear at 79, 63, 73, 72%. They are mostly consumed by (local storage) kube VMs.
- Do we have any cold spares ready?
- How does the replacement work? Wait for 99% wearout? Offline dd? Half-online zfs mirror?
***
- Kube local VMs seem to not be able to migrate to other nodes.
- Kindly stop requiring local storage for kube applications :)
***
- This should be automatically tracked.
rasmus
added this to the k-space.ee/todo project 2025-02-15 20:10:01 +00:00
2025-04-05: 81 65 75 74 (+2% on each). Napking math projection says about a year until there really needs to be orders for replacements.
2025-05-24: 83 68 77 75
2025-06-28: 85 70 79 76
2025-07-21: 87 71 80 77
2025-04-05: 81 65 75 74 (+2% on each). Napking math projection says about a year until there really needs to be orders for replacements.
2025-05-24: 83 68 77 75
2025-06-28: 85 70 79 76
2025-07-21: 87 71 80 77
rasmus
added the due date 2026-03-152025-04-05 20:02:04 +00:00
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
The four Proxmox nodes have one (SAMSUNG MZ7LH480HAHQ 480G) SSD each with disk wear at 79, 63, 73, 72%. They are mostly consumed by (local storage) kube VMs.
2025-04-05: 81 65 75 74 (+2% on each). Napking math projection says about a year until there really needs to be orders for replacements.
2025-05-24: 83 68 77 75
2025-06-28: 85 70 79 76
2025-07-21: 87 71 80 77
Probably would make sense to replace those drives with newer nvme drives during blade -> hp gen9 pizza box migration
PVE has been migrated over to CEPH and NVME. This is no longer relevant. Need to create a new issue for future drives.