: 31.0 GiBSwap : 1024.0 MiBDisk : 418.3 GiBDistro : AlmaLinux 10.1 (Heliotrope Lion)Kernel : 6.12.0-124.45.1.el10_1.x86_64VM Type : NONEIPv4/IPv6 : ✔ Online / ✔ OnlineI

: 31.2 GiBSwap : 1024.0 MiBDisk : 446.2 GiBDistro : AlmaLinux 9.7 (Moss Jungle Cat)Kernel : 5.14.0-611.42.1.el9_7.x86_64VM Type : NONEIPv4/IPv6 : ✔ Online / ✔ OnlineIPv

(Quote)

DC: gra1CPU: E3-1245v2RAM: 32768 MBDISKS: 2 X Disk SSD 480 GB, JBOD

(Quote)

Hm, mine Dus storage-1 has "tsc" by default (and I have 1G upload speed)

I see it's trying to boot but it reports "no bootable disk". You did not change anything in the partition scheme/made bootloader upgrade/changes, right?

After the initial deployment, clocksource was set to hpet by the cloud-init script. A hard reboot via the control panel switched it back to kvm-clock, and upload speed jumped immediately.

I also saw that at first, after I upgraded it to Debian trixie, the speeds increased significantly, so at first I believed it had something to do with a newer kernel or similar.

@emperor Thanks for the tip! Reran YABS from the data disk (LUKS2 + btrfs on the HDD array):

KS-2 or KS-6 with 10Gbit upload speed - Can pay $$ or trade following servers:

Run yabs fromd /data (or where ever you mounted the disk) and will do it from hdd instead ssd. The script uses current directory for disk test.

(Quote)

(Quote)

(Quote)

: ✔ Enabled VM-x/AMD-V : ✔ Enabled Total Disk : 24.0 GB (2.1 GB Used) Total RAM : 1.9 GB (210.4 MB Used) Total Swap : 512.0 MB (0 Used) System uptime : 0 days, 0

* Disk: 10G SATA SSD RAID10 (IO Limited)

EnabledRAM : 1.9 GiBSwap : 512.0 MiBDisk : 24.0 GiBDistro : Debian GNU/Linux 12 (bookworm)Kernel : 6.1.0-44-amd64VM Type : KVMIPv4/IPv6 : ✔ Online / ❌ OfflineIPv4 Netwo

Notes: The "19.5 TiB" disk figure is inflated — YABS sums all mount points, and btrfs subvolumes each report the full 4.9 TB, so it gets counted 4x. Actual storage is 10 GB system SS

(Quote)