I am troubleshooting a disk that has realtively horrible performace and I am curious... does Proxmox Backup write using sync or async? As if it's sync, I probably need a separate ZIL or to forget about ZFS. If async, well... I keep digging.
Writes sync or async?
- Thread starter Bubbagump210
- Start date
I assume you're talking about PBS - the server side?
I just checked that by running strace - the writes happen to files opened as:
(and the writes themselves don't issue syncs)
The following is not recommended for production use and should only be set temporary for testing!!!!:
in any case if you want to try if a slog/zil device would help you could try setting the sync-option of the dataset to off
(if this improves the performance - adding a fast slog device should also improve the situation (though not as much)
set the setting back to default after you did your tests!
I hope this helps!
I just checked that by running strace - the writes happen to files opened as:
O_WRONLY|O_CREAT|O_TRUNC|O_CLOEXEC(and the writes themselves don't issue syncs)
The following is not recommended for production use and should only be set temporary for testing!!!!:
in any case if you want to try if a slog/zil device would help you could try setting the sync-option of the dataset to off
(if this improves the performance - adding a fast slog device should also improve the situation (though not as much)
set the setting back to default after you did your tests!
I hope this helps!
Yes indeed, I am talking about the server side. You lost me though.
leads me to believe writes are async by default? Are you saying that writes are synchronous or no as then you say
which makes me think writes ARE synchronous? And how would I enable/disable synchronous writes? Thanks!
Last edited:
the term async is quite broad - so I wasn't 100% sure what you mean (hence the resulting confusion
- in the context of the zfs property sync (and in the context of adding slog devices) the writes are not sync (the file is not opened with O_DSYNC, and we don't see calls to fsync(2).no - I tried to point to a way of finding out if adding a SLOG/ZIL device would improve the performance of the datastore (without simply adding a slog device)
this is not configurable in proxmox-backup - but you can try the effect by modifying the 'sync' property (see `man zfs` (or if you've updated to zfs > 2.0.0 `man zfsprops`)
I hope this helps!
That’s what I was after. No fsync() means async in my book. I suspect a SLOG won’t help at all then. I’ll try disabling the ZFS property via /sys and see if it changes anything, but yeah, I think this disk is just not up to the task and it’s not a double write ZIL issue. Thanks!
I suspect a SLOG won’t help at all then. I’ll try disabling the ZFS property via /sys and see if it changes anything, but yeah, I think this disk is just not up to the task and it’s not a double write ZIL issue. Thanks!This is quite old, but did you managed to increase performance somehow?
Im asking myself all the same questions xD
But i disabled sync on the pool and i see no performance benefits in speed, so i think adding a slog device wont help.
Thats a Genoa Server with 9374f, 12 memory channels populated (768GB) and 8x Micron 7450 in Raid10.
The Connection to the Backup Server is 2x25G in LACP.
The Backupserver itself has 2x E5-2637 v3 and 6x 14TB SAS-HDDs in Raid 10 and another pool with a stripe of 5x Raid-Z1(3x4TB) SATA Drives.
So totally 21 Drives, connected via LSI 9305-24i (No Multiplexer Card)
And each CPU on the Backupserver is populated with all 4 Memory Channels (256gb in total)
The Network Card sits on CPU 1 and the Storage Card sits on CPU 2, so interconnect might be an issue, but interconnect is surely not limited to ~200mb xD
The Backupserver does absolutely nothing else as playing PBS.
And the Genoa Server is single Socket, so no interconnect there.
But both pools on the backupserver have deduplication enabled, that costs surely some speed, but the memory bandwith is still fast. I mean that also shouldn't limit to 200mb/s.
So my issue is, that i have Absolutely no Clue what i could do to speed it up. And from the Backup logs, it is looking to me tbh, like the Genoa Server cant deliver more, its like throttling the read speed from the SSD's. But that are 8x Micron7450 Max.
I did even nvme format to set the blocksize to 4kb native on them, because they come out of the box 512b formatted.
I could set MTU to 9000, but that would require to insert another Network Card into all Servers, because the Actual Network Cards have a lot of Vlans etc on them, need there 1500. However, i doubt highly that even this would bring more performance, because we talk about 200mb/s speeds...
Not 5Gb/s speeds, where MTU might start to play a role.
If anyone has ideas, suggestions, you're welcome!
Cheers
Im asking myself all the same questions xD
But i disabled sync on the pool and i see no performance benefits in speed, so i think adding a slog device wont help.
Code:
INFO: starting new backup job: vzdump 138 --node pve-hdr --notes-template '{{guestname}}' --remove 0 --storage Backup-SAS --notification-mode auto --mode snapshot
INFO: Starting Backup of VM 138 (qemu)
INFO: Backup started at 2024-05-03 18:13:27
INFO: status = running
INFO: VM Name: DokuBit
INFO: include disk 'virtio0' 'Storage-Default:vm-138-disk-0' 200G
INFO: include disk 'virtio1' 'Storage-Default:vm-138-disk-1' 150G
INFO: include disk 'virtio2' 'Storage-Default:vm-138-disk-2' 1500G
INFO: include disk 'virtio3' 'Storage-Default:vm-138-disk-3' 200G
INFO: backup mode: snapshot
INFO: ionice priority: 7
INFO: creating Proxmox Backup Server archive 'vm/138/2024-05-03T16:13:27Z'
INFO: issuing guest-agent 'fs-freeze' command
INFO: issuing guest-agent 'fs-thaw' command
INFO: started backup task 'af2454a8-e65f-4e2e-a6d1-8268d2b3c595'
INFO: resuming VM again
INFO: virtio0: dirty-bitmap status: OK (4.9 GiB of 200.0 GiB dirty)
INFO: virtio1: dirty-bitmap status: OK (12.8 GiB of 150.0 GiB dirty)
INFO: virtio2: dirty-bitmap status: OK (7.0 GiB of 1.5 TiB dirty)
INFO: virtio3: dirty-bitmap status: OK (10.3 GiB of 200.0 GiB dirty)
INFO: using fast incremental mode (dirty-bitmap), 35.1 GiB dirty of 2.0 TiB total
INFO: 4% (1.5 GiB of 35.1 GiB) in 3s, read: 504.0 MiB/s, write: 504.0 MiB/s
INFO: 6% (2.4 GiB of 35.1 GiB) in 6s, read: 302.7 MiB/s, write: 302.7 MiB/s
INFO: 9% (3.4 GiB of 35.1 GiB) in 9s, read: 338.7 MiB/s, write: 338.7 MiB/s
INFO: 11% (4.0 GiB of 35.1 GiB) in 12s, read: 222.7 MiB/s, write: 222.7 MiB/s
INFO: 12% (4.3 GiB of 35.1 GiB) in 15s, read: 84.0 MiB/s, write: 84.0 MiB/s
INFO: 13% (4.7 GiB of 35.1 GiB) in 22s, read: 70.3 MiB/s, write: 70.3 MiB/s
INFO: 14% (4.9 GiB of 35.1 GiB) in 25s, read: 62.7 MiB/s, write: 62.7 MiB/s
INFO: 15% (5.3 GiB of 35.1 GiB) in 30s, read: 73.6 MiB/s, write: 73.6 MiB/s
INFO: 16% (5.6 GiB of 35.1 GiB) in 36s, read: 60.0 MiB/s, write: 60.0 MiB/s
INFO: 17% (6.0 GiB of 35.1 GiB) in 41s, read: 70.4 MiB/s, write: 70.4 MiB/s
INFO: 18% (6.3 GiB of 35.1 GiB) in 46s, read: 71.2 MiB/s, write: 71.2 MiB/s
INFO: 19% (6.7 GiB of 35.1 GiB) in 51s, read: 85.6 MiB/s, write: 85.6 MiB/s
INFO: 20% (7.1 GiB of 35.1 GiB) in 55s, read: 85.0 MiB/s, write: 85.0 MiB/s
INFO: 21% (7.4 GiB of 35.1 GiB) in 1m, read: 66.4 MiB/s, write: 66.4 MiB/s
INFO: 22% (7.8 GiB of 35.1 GiB) in 1m 5s, read: 82.4 MiB/s, write: 82.4 MiB/s
INFO: 23% (8.1 GiB of 35.1 GiB) in 1m 8s, read: 97.3 MiB/s, write: 97.3 MiB/s
INFO: 24% (8.5 GiB of 35.1 GiB) in 1m 13s, read: 76.0 MiB/s, write: 76.0 MiB/s
INFO: 25% (8.9 GiB of 35.1 GiB) in 1m 23s, read: 46.4 MiB/s, write: 46.4 MiB/s
INFO: 26% (9.5 GiB of 35.1 GiB) in 1m 26s, read: 188.0 MiB/s, write: 188.0 MiB/s
INFO: 27% (9.7 GiB of 35.1 GiB) in 1m 29s, read: 78.7 MiB/s, write: 78.7 MiB/s
INFO: 28% (10.0 GiB of 35.1 GiB) in 1m 36s, read: 45.7 MiB/s, write: 45.7 MiB/s
INFO: 29% (10.5 GiB of 35.1 GiB) in 1m 39s, read: 174.7 MiB/s, write: 174.7 MiB/s
INFO: 30% (10.8 GiB of 35.1 GiB) in 1m 42s, read: 92.0 MiB/s, write: 92.0 MiB/s
INFO: 31% (11.2 GiB of 35.1 GiB) in 1m 46s, read: 109.0 MiB/s, write: 108.0 MiB/s
INFO: 32% (11.2 GiB of 35.1 GiB) in 1m 49s, read: 10.7 MiB/s, write: 9.3 MiB/s
INFO: 33% (11.7 GiB of 35.1 GiB) in 1m 54s, read: 87.2 MiB/s, write: 86.4 MiB/s
INFO: 34% (12.0 GiB of 35.1 GiB) in 1m 58s, read: 83.0 MiB/s, write: 83.0 MiB/s
INFO: 35% (12.3 GiB of 35.1 GiB) in 2m 2s, read: 79.0 MiB/s, write: 79.0 MiB/s
INFO: 36% (12.6 GiB of 35.1 GiB) in 2m 7s, read: 68.8 MiB/s, write: 68.8 MiB/s
INFO: 37% (13.0 GiB of 35.1 GiB) in 2m 12s, read: 80.8 MiB/s, write: 80.8 MiB/s
INFO: 38% (13.4 GiB of 35.1 GiB) in 2m 21s, read: 45.8 MiB/s, write: 45.8 MiB/s
INFO: 39% (14.0 GiB of 35.1 GiB) in 2m 24s, read: 184.0 MiB/s, write: 184.0 MiB/s
INFO: 40% (14.2 GiB of 35.1 GiB) in 2m 27s, read: 66.7 MiB/s, write: 66.7 MiB/s
INFO: 41% (14.4 GiB of 35.1 GiB) in 2m 30s, read: 76.0 MiB/s, write: 76.0 MiB/s
INFO: 42% (14.8 GiB of 35.1 GiB) in 2m 35s, read: 83.2 MiB/s, write: 82.4 MiB/s
INFO: 43% (15.1 GiB of 35.1 GiB) in 2m 39s, read: 77.0 MiB/s, write: 76.0 MiB/s
INFO: 44% (15.5 GiB of 35.1 GiB) in 2m 44s, read: 78.4 MiB/s, write: 78.4 MiB/s
INFO: 45% (15.8 GiB of 35.1 GiB) in 2m 48s, read: 85.0 MiB/s, write: 84.0 MiB/s
INFO: 46% (16.2 GiB of 35.1 GiB) in 2m 52s, read: 94.0 MiB/s, write: 94.0 MiB/s
INFO: 47% (16.5 GiB of 35.1 GiB) in 2m 56s, read: 93.0 MiB/s, write: 93.0 MiB/s
INFO: 48% (16.9 GiB of 35.1 GiB) in 3m, read: 85.0 MiB/s, write: 85.0 MiB/s
INFO: 49% (17.2 GiB of 35.1 GiB) in 3m 5s, read: 72.0 MiB/s, write: 72.0 MiB/s
INFO: 50% (17.6 GiB of 35.1 GiB) in 3m 10s, read: 72.8 MiB/s, write: 72.8 MiB/s
INFO: 51% (18.2 GiB of 35.1 GiB) in 3m 14s, read: 166.0 MiB/s, write: 166.0 MiB/s
INFO: 52% (18.3 GiB of 35.1 GiB) in 3m 19s, read: 22.4 MiB/s, write: 22.4 MiB/s
INFO: 53% (18.7 GiB of 35.1 GiB) in 3m 22s, read: 122.7 MiB/s, write: 122.7 MiB/s
INFO: 54% (19.0 GiB of 35.1 GiB) in 3m 26s, read: 67.0 MiB/s, write: 67.0 MiB/s
INFO: 55% (19.3 GiB of 35.1 GiB) in 3m 31s, read: 76.8 MiB/s, write: 76.8 MiB/s
INFO: 56% (19.7 GiB of 35.1 GiB) in 3m 35s, read: 88.0 MiB/s, write: 88.0 MiB/s
INFO: 57% (20.0 GiB of 35.1 GiB) in 3m 39s, read: 82.0 MiB/s, write: 82.0 MiB/s
INFO: 58% (20.4 GiB of 35.1 GiB) in 3m 44s, read: 77.6 MiB/s, write: 77.6 MiB/s
INFO: 59% (20.7 GiB of 35.1 GiB) in 3m 47s, read: 125.3 MiB/s, write: 125.3 MiB/s
INFO: 60% (21.1 GiB of 35.1 GiB) in 3m 51s, read: 95.0 MiB/s, write: 95.0 MiB/s
INFO: 61% (21.4 GiB of 35.1 GiB) in 3m 55s, read: 76.0 MiB/s, write: 76.0 MiB/s
INFO: 62% (21.8 GiB of 35.1 GiB) in 3m 59s, read: 92.0 MiB/s, write: 92.0 MiB/s
INFO: 63% (22.1 GiB of 35.1 GiB) in 4m 2s, read: 112.0 MiB/s, write: 112.0 MiB/s
INFO: 64% (22.6 GiB of 35.1 GiB) in 4m 6s, read: 129.0 MiB/s, write: 129.0 MiB/s
INFO: 66% (23.2 GiB of 35.1 GiB) in 4m 11s, read: 116.0 MiB/s, write: 116.0 MiB/s
INFO: 67% (23.6 GiB of 35.1 GiB) in 4m 15s, read: 113.0 MiB/s, write: 113.0 MiB/s
INFO: 68% (24.2 GiB of 35.1 GiB) in 4m 18s, read: 188.0 MiB/s, write: 188.0 MiB/s
INFO: 69% (24.5 GiB of 35.1 GiB) in 4m 21s, read: 106.7 MiB/s, write: 106.7 MiB/s
INFO: 70% (24.9 GiB of 35.1 GiB) in 4m 24s, read: 133.3 MiB/s, write: 133.3 MiB/s
INFO: 72% (25.3 GiB of 35.1 GiB) in 4m 27s, read: 149.3 MiB/s, write: 149.3 MiB/s
INFO: 73% (25.7 GiB of 35.1 GiB) in 4m 32s, read: 81.6 MiB/s, write: 81.6 MiB/s
INFO: 74% (26.1 GiB of 35.1 GiB) in 4m 35s, read: 126.7 MiB/s, write: 126.7 MiB/s
INFO: 76% (26.8 GiB of 35.1 GiB) in 4m 38s, read: 237.3 MiB/s, write: 237.3 MiB/s
INFO: 77% (27.1 GiB of 35.1 GiB) in 4m 41s, read: 120.0 MiB/s, write: 120.0 MiB/s
INFO: 78% (27.5 GiB of 35.1 GiB) in 4m 44s, read: 138.7 MiB/s, write: 138.7 MiB/s
INFO: 79% (28.0 GiB of 35.1 GiB) in 4m 47s, read: 146.7 MiB/s, write: 146.7 MiB/s
INFO: 80% (28.3 GiB of 35.1 GiB) in 4m 50s, read: 113.3 MiB/s, write: 113.3 MiB/s
INFO: 81% (28.7 GiB of 35.1 GiB) in 4m 53s, read: 145.3 MiB/s, write: 145.3 MiB/s
INFO: 83% (29.1 GiB of 35.1 GiB) in 4m 56s, read: 140.0 MiB/s, write: 140.0 MiB/s
INFO: 84% (29.6 GiB of 35.1 GiB) in 4m 59s, read: 149.3 MiB/s, write: 149.3 MiB/s
INFO: 85% (29.9 GiB of 35.1 GiB) in 5m 2s, read: 126.7 MiB/s, write: 126.7 MiB/s
INFO: 86% (30.3 GiB of 35.1 GiB) in 5m 5s, read: 128.0 MiB/s, write: 128.0 MiB/s
INFO: 87% (30.8 GiB of 35.1 GiB) in 5m 8s, read: 177.3 MiB/s, write: 177.3 MiB/s
INFO: 88% (31.1 GiB of 35.1 GiB) in 5m 11s, read: 108.0 MiB/s, write: 108.0 MiB/s
INFO: 89% (31.5 GiB of 35.1 GiB) in 5m 14s, read: 137.3 MiB/s, write: 134.7 MiB/s
INFO: 90% (31.8 GiB of 35.1 GiB) in 5m 17s, read: 89.3 MiB/s, write: 89.3 MiB/s
INFO: 91% (32.1 GiB of 35.1 GiB) in 5m 20s, read: 112.0 MiB/s, write: 109.3 MiB/s
INFO: 92% (32.4 GiB of 35.1 GiB) in 5m 23s, read: 97.3 MiB/s, write: 93.3 MiB/s
INFO: 93% (32.7 GiB of 35.1 GiB) in 5m 26s, read: 89.3 MiB/s, write: 89.3 MiB/s
INFO: 94% (33.0 GiB of 35.1 GiB) in 5m 30s, read: 87.0 MiB/s, write: 87.0 MiB/s
INFO: 95% (33.3 GiB of 35.1 GiB) in 5m 34s, read: 80.0 MiB/s, write: 80.0 MiB/s
INFO: 96% (33.7 GiB of 35.1 GiB) in 5m 38s, read: 102.0 MiB/s, write: 101.0 MiB/s
INFO: 97% (34.3 GiB of 35.1 GiB) in 5m 41s, read: 181.3 MiB/s, write: 180.0 MiB/s
INFO: 98% (34.6 GiB of 35.1 GiB) in 5m 44s, read: 116.0 MiB/s, write: 116.0 MiB/s
INFO: 99% (35.0 GiB of 35.1 GiB) in 5m 47s, read: 137.3 MiB/s, write: 134.7 MiB/s
INFO: 100% (35.1 GiB of 35.1 GiB) in 5m 50s, read: 24.0 MiB/s, write: 22.7 MiB/s
INFO: backup was done incrementally, reused 1.97 TiB (98%)
INFO: transferred 35.07 GiB in 357 seconds (100.6 MiB/s)
INFO: adding notes to backup
INFO: Finished Backup of VM 138 (00:05:59)
INFO: Backup finished at 2024-05-03 18:19:26
INFO: Backup job finished successfully
INFO: notified via target `mail-to-root`
TASK OKThats a Genoa Server with 9374f, 12 memory channels populated (768GB) and 8x Micron 7450 in Raid10.
The Connection to the Backup Server is 2x25G in LACP.
The Backupserver itself has 2x E5-2637 v3 and 6x 14TB SAS-HDDs in Raid 10 and another pool with a stripe of 5x Raid-Z1(3x4TB) SATA Drives.
So totally 21 Drives, connected via LSI 9305-24i (No Multiplexer Card)
And each CPU on the Backupserver is populated with all 4 Memory Channels (256gb in total)
The Network Card sits on CPU 1 and the Storage Card sits on CPU 2, so interconnect might be an issue, but interconnect is surely not limited to ~200mb xD
The Backupserver does absolutely nothing else as playing PBS.
And the Genoa Server is single Socket, so no interconnect there.
But both pools on the backupserver have deduplication enabled, that costs surely some speed, but the memory bandwith is still fast. I mean that also shouldn't limit to 200mb/s.
So my issue is, that i have Absolutely no Clue what i could do to speed it up. And from the Backup logs, it is looking to me tbh, like the Genoa Server cant deliver more, its like throttling the read speed from the SSD's. But that are 8x Micron7450 Max.
I did even nvme format to set the blocksize to 4kb native on them, because they come out of the box 512b formatted.
I could set MTU to 9000, but that would require to insert another Network Card into all Servers, because the Actual Network Cards have a lot of Vlans etc on them, need there 1500. However, i doubt highly that even this would bring more performance, because we talk about 200mb/s speeds...
Not 5Gb/s speeds, where MTU might start to play a role.
If anyone has ideas, suggestions, you're welcome!
Cheers
Last edited:
Code:
INFO: starting new backup job: vzdump 150 --storage Backup-SAS --notes-template '{{guestname}}' --remove 0 --mode snapshot --notification-mode auto --node pve-bdr
INFO: Starting Backup of VM 150 (qemu)
INFO: Backup started at 2024-05-03 19:29:43
INFO: status = running
INFO: VM Name: DokuBit2
INFO: include disk 'virtio0' 'Storage-Default:vm-150-disk-0' 100G
INFO: include disk 'virtio1' 'Storage-Default:vm-150-disk-1' 1T
INFO: include disk 'virtio2' 'Storage-Default:vm-150-disk-2' 1T
INFO: include disk 'virtio3' 'Storage-Default:vm-150-disk-3' 1T
INFO: backup mode: snapshot
INFO: ionice priority: 7
INFO: creating Proxmox Backup Server archive 'vm/150/2024-05-03T17:29:43Z'
INFO: issuing guest-agent 'fs-freeze' command
INFO: issuing guest-agent 'fs-thaw' command
INFO: started backup task '3fdfd40a-c89a-4adf-bda7-558bfcd26a42'
INFO: resuming VM again
INFO: virtio0: dirty-bitmap status: OK (2.7 GiB of 100.0 GiB dirty)
INFO: virtio1: dirty-bitmap status: OK (44.5 GiB of 1.0 TiB dirty)
INFO: virtio2: dirty-bitmap status: OK (7.5 GiB of 1.0 TiB dirty)
INFO: virtio3: dirty-bitmap status: OK (972.0 MiB of 1.0 TiB dirty)
INFO: using fast incremental mode (dirty-bitmap), 55.7 GiB dirty of 3.1 TiB total
INFO: 3% (1.9 GiB of 55.7 GiB) in 3s, read: 654.7 MiB/s, write: 648.0 MiB/s
INFO: 6% (3.7 GiB of 55.7 GiB) in 6s, read: 597.3 MiB/s, write: 593.3 MiB/s
INFO: 9% (5.1 GiB of 55.7 GiB) in 9s, read: 480.0 MiB/s, write: 480.0 MiB/s
INFO: 13% (7.4 GiB of 55.7 GiB) in 12s, read: 810.7 MiB/s, write: 805.3 MiB/s
INFO: 17% (9.6 GiB of 55.7 GiB) in 15s, read: 728.0 MiB/s, write: 718.7 MiB/s
INFO: 22% (12.4 GiB of 55.7 GiB) in 18s, read: 970.7 MiB/s, write: 786.7 MiB/s
INFO: 26% (14.9 GiB of 55.7 GiB) in 21s, read: 841.3 MiB/s, write: 841.3 MiB/s
INFO: 31% (17.5 GiB of 55.7 GiB) in 24s, read: 900.0 MiB/s, write: 900.0 MiB/s
INFO: 36% (20.1 GiB of 55.7 GiB) in 27s, read: 894.7 MiB/s, write: 894.7 MiB/s
INFO: 40% (22.6 GiB of 55.7 GiB) in 30s, read: 826.7 MiB/s, write: 826.7 MiB/s
INFO: 45% (25.3 GiB of 55.7 GiB) in 33s, read: 928.0 MiB/s, write: 928.0 MiB/s
INFO: 50% (28.0 GiB of 55.7 GiB) in 36s, read: 929.3 MiB/s, write: 929.3 MiB/s
INFO: 55% (30.7 GiB of 55.7 GiB) in 39s, read: 932.0 MiB/s, write: 932.0 MiB/s
INFO: 59% (33.3 GiB of 55.7 GiB) in 42s, read: 885.3 MiB/s, write: 885.3 MiB/s
INFO: 64% (36.1 GiB of 55.7 GiB) in 45s, read: 944.0 MiB/s, write: 944.0 MiB/s
INFO: 69% (38.9 GiB of 55.7 GiB) in 48s, read: 954.7 MiB/s, write: 954.7 MiB/s
INFO: 74% (41.5 GiB of 55.7 GiB) in 51s, read: 890.7 MiB/s, write: 890.7 MiB/s
INFO: 79% (44.1 GiB of 55.7 GiB) in 54s, read: 876.0 MiB/s, write: 876.0 MiB/s
INFO: 84% (46.9 GiB of 55.7 GiB) in 57s, read: 957.3 MiB/s, write: 957.3 MiB/s
INFO: 89% (50.0 GiB of 55.7 GiB) in 1m, read: 1.1 GiB/s, write: 1.1 GiB/s
INFO: 93% (52.3 GiB of 55.7 GiB) in 1m 3s, read: 778.7 MiB/s, write: 776.0 MiB/s
INFO: 97% (54.1 GiB of 55.7 GiB) in 1m 6s, read: 600.0 MiB/s, write: 594.7 MiB/s
INFO: 99% (55.3 GiB of 55.7 GiB) in 1m 9s, read: 426.7 MiB/s, write: 425.3 MiB/s
INFO: 100% (55.7 GiB of 55.7 GiB) in 1m 12s, read: 141.3 MiB/s, write: 141.3 MiB/s
INFO: Waiting for server to finish backup validation...
INFO: backup was done incrementally, reused 3.04 TiB (98%)
INFO: transferred 55.75 GiB in 95 seconds (600.9 MiB/s)
INFO: adding notes to backup
INFO: Finished Backup of VM 150 (00:01:39)
INFO: Backup finished at 2024-05-03 19:31:22
INFO: Backup job finished successfully
INFO: notified via target `mail-to-root`
TASK OKI solved the issue myself, it wasn't a problem of deduplication, glad god
Im getting now speeds near the limits of my 6x 14TB Raid10 Pool. The other Pool still needs to be tested, but i think its the same, i just never used the other pool anyway, but it should be even faster.
However all it was, was recordsize=1m, xattr=sa, dnodesize=auto and logbias=throughput
It was all defaults before, i forgot about those settings tbh. But the difference is huge, so glad god i don't need to disable deduplication.
Cheers
Last edited: