Wasabi wallet (2.8.1:2) backup on a fresh 0.4.0.1 install is taking a ridiculously long time. It’s been 8h+ and it is only at 82%.
On the old OS version, backups were done in minutes.
I can’t see anything unusual in the logs.
USB drives are 64GB and new, so that’s not the issue.
Anyone else having the same experience?
We need a cancel button in backups to retry the backup without the service that’s stuck, because some services can take way too long to back up (Ollama, for example), and we need to cancel it safely without needing to unplug the hardware for backups.
Hey @stellarstoic, sorry about the issue you’re experiencing. Couple of quick questions:
Is this an initial backup or incremental? If this is a wallet with a large UTXO set / big SQLite or LMDB-backed datadir, and the backup mechanism is doing a full copy rather than anything incremental, that’s a plausible source of slowness, especially if there’s fsync-heavy behavior during the copy.
82% stuck for hours vs progressing slowly - is the percentage actually moving at all, just slowly, or is it flatlined at 82%? Those are different failure modes. Flatlined for hours strongly suggests something is blocked (I/O wait, a lock, a hung subprocess) rather than “just slow.”
Yes this wallet I imported does have a large UTXO set and it is an initial backup.
the percentages are moving but realy slow. It just finished the backup finally.
I tried to trigger backup again with Hermes and Wasabi and Hermes looks like it’s stuck at 100%(it did finished after an hour) and the Wasabi again is slugish as hell.
I think I simply create a new wasabi wallet.
I did noticed another service the other day, Ollama that was stuck at 100% for hours but I dared to unplug the USB dongle and start again without ollama.
Hi @stellarstoic thanks for the update. If you’re willing, next time a backup is sitting at 100% for a while, could you SSH in and run iotop (or just check dmesg for anything recent) to see whether the drive is still actively being written to, or whether it’s gone quiet? That tells us if it’s genuinely still flushing data or actually hung.
Long wasabi backup because of the large UTXO set CONFIRMED.
It took not even a minute to backup the new wallet.
I also retried the other services that previously stuck at 100%(Not sure if they actually stucked. they just show 100% for a long time).
I backed up one by one not all at the same time and that actually went smoothly. no long time stuck on 100% anymore. Maybe there’s some glitch when we backing up a large batch of services at once idk.
If something stucks at 100% again I will ssh in and give you the report for the dmesg. iotop is too noisy.
