#duplicacy — Public Fediverse posts

I’m thinking that rsync is not sufficient for important large data backups. Even if I’m rsyncing the data away to the Hetzner Storage Box.
I need something more advanced with deduplication and so on. I tried #borgbackup in the past, and that thing was absolutely not user-friendly. It was rather user-hostile.
Now, #SelfHosted community has #Duplicati and #Duplicacy? Does someone use one? Any recommendations?

#homelab #selfhosting #selfhost #backup #backups #databackup

I’m thinking that rsync is not sufficient for important large data backups. Even if I’m rsyncing the data away to the Hetzner Storage Box.
I need something more advanced with deduplication and so on. I tried #borgbackup in the past, and that thing was absolutely not user-friendly. It was rather user-hostile.
Now, #SelfHosted community has #Duplicati and #Duplicacy? Does someone use one? Any recommendations?

#homelab #selfhosting #selfhost #backup #backups #databackup

Acrosync LLC released #Duplicacy CLI version 3.2.5. https://duplicacy.com/download.html

Acrosync LLC released #Duplicacy CLI version 3.2.5. https://duplicacy.com/download.html

I'm still new to trying out large offsite cloud backups, but using #Duplicacy to #BackBlaze B2, it backed up 121GB of data over T-Mobile 5G over the course of < 10.5 hours.

Speed-wise it felt slow, but compared to the "big comparison" benchmark it wasn't too bad.

I'm still new to trying out large offsite cloud backups, but using #Duplicacy to #BackBlaze B2, it backed up 121GB of data over T-Mobile 5G over the course of < 10.5 hours.

Speed-wise it felt slow, but compared to the "big comparison" benchmark it wasn't too bad.

@kielbasa It is indeed! I use #duplicacy running under #docker on my backup server, with NFS mounts of the host machine and my primary #unraid server. It periodically makes backups both to its local array as well as off-site to #backblaze

I'm transitioning away from it, but I also have a #proxmox cluster that is backed up to the same box using Proxmox's backup server. So it's actually responsible for backing up quite a lot.

Outside of this I also occasionally burn my most precious files to UHD bluray and store those off-site as well.

@kielbasa It is indeed! I use #duplicacy running under #docker on my backup server, with NFS mounts of the host machine and my primary #unraid server. It periodically makes backups both to its local array as well as off-site to #backblaze

I'm transitioning away from it, but I also have a #proxmox cluster that is backed up to the same box using Proxmox's backup server. So it's actually responsible for backing up quite a lot.

Outside of this I also occasionally burn my most precious files to UHD bluray and store those off-site as well.

@rune I'm using #duplicacy, sharing repos between clients has never been a problem. Compression and deduplication work well, and I never had any corruption issues. The project had been a bit slow in the last couple of years, but it's not abandoned

создание инкиментального архива(duplicacy how-to)

Я тут пересмотрел современные бэкапилки. Выбор пал на duplicacy (не путать с duplicati и duplicity). Выбранный бесплатен для персонального использования. Платный для коммерческого, но копейки: $50 за тазик в год. Хотя cli версия опенсорсная и лицензция нигде не проверяется. В основном современные бэкапилки с дедубликацией очень похожи между собой, делают одно и то же одинаковыми библиотеками. Самый фичастый с большим функционалом restic, но в нём нет сжатия. Со сжатием аналогичный borg, но из коробки не умеет лить в облака. Кстати, те вариант, которые льют в облака - не хранят локальную копию. Ещё особняком стоит duplicati. Он фичастый, крутой, обоссаться прям. Есть реально всё, что можно придумать. Но на дотнете 🤦‍♂️ Вроде есть пакеты под всё кроме фряхи, но фиг знает, недоверительно. Хотя мне разок попалась компания, у которой прод на дотнете под линуксом на моно. А duplicacy выбрал, в основном из-за этого: -keep 0:777 -keep 33:386 -keep 22:144 -keep 9:44 -keep 3:11 -keep 1:1 Keep no snapshots older than 777 days Keep 1 snapshot every 33 day(s) if older than 386 day(s) Keep 1 snapshot every 22 day(s) if older than 144 day(s) Keep 1 snapshot every 9 day(s) if older than 44 day(s) Keep 1 snapshot every 3 day(s) if older than 11 day(s) Keep 1 snapshot every 1 day(s) if older than 1 day(s) Такое есть только у duplicati (который на дотнете). У остальных только день, месяц, год. Т.е. произвольный retention задать нельзя. Хотя по фичастости duplicacy довольно скромен. Скромнее того же restic или borg. 1) скачиваем последний релиз 2) переходим к папке, которую нужно резервировать и инициализируем место хранения duplicacy init [command options] "snapshot id" "storage url" duplicacy init -zstd -storage-name yandex-disk id_1 /app/yandex-disk/ ... /app/www/html/domostroy/files/image will be backed up to /app/yandex-disk/ with id id_1 3) инкриментальное резервное копирование в папке осуществляется командой duplicacy backup 4) просмотр истории файла в архиве осуществляется командой duplicacy history 5) восстановление файла duplicacy restore [command options] [--] [pattern] ... duplicacy restore -r 3 [filename] восстанавивает и директорию файла Просмотр снепшотов в репозитории duplicacy list удаление ненужных снапшотов duplicacy prune -r 6 чтобы найти файл в архиве duplicacy cat | grep <имя файла> Читать далее почему Выбор пал на duplicacy

https://habr.com/ru/articles/792574/

#duplicacy

Acrosync LLC released #Duplicacy CLI version 3.2.3. https://duplicacy.com/download.html

Acrosync LLC released #Duplicacy CLI version 3.2.2. https://duplicacy.com/download.html

Acrosync LLC released #Duplicacy CLI version 3.2.0. https://duplicacy.com/download.html

Acrosync LLC released #Duplicacy CLI version 3.1.0. https://duplicacy.com/download.html

@urusan note that syncthing is not a #backup solution. It has the capability to have file versions, but (for example) you cannot mount or browse time based snapshots etc, and file deletion or corruption in the source will get propagated to the sync target.

For backup, I've used #Duplicacy and #Kopia in the past with good results.

If you want a live browsable copy on the Linux system with two way sync, then you will need syncthing.

@publicvoit
Mhh #restic war mir noch gar nicht bekannt. Nach einigen Vergleichen hab ich derzeit #duplicacy und #rsync im Einsatz, je nach BackupDst.