size versus usage mismatch

[dejonghb]

As someone already claimed: screenshot says it all... (hint: size versus usage)

[kvanhijf]

Values come from alba

[dejonghb]

root@n19:~# alba dev-extract-config --config arakoon://config/ovs/alba/asds/R91eDx61ZQqXegDmJKvSVdCHvowGek2J/config?ini=%2Fopt%2Fasd-manager%2Fconfig%2Farakoon_cacc.ini
{"asd_id": "R91eDx61ZQqXegDmJKvSVdCHvowGek2J", "capacity": 202451916800, "multicast": null, "log_level": "info", "ips": ["172.22.186.119"], "node_id": "tcKGJSiZAL4q99qJ", "transport": "tcp", "home": "/mnt/alba-asd/n19-fhdd3/R91eDx61ZQqXegDmJKvSVdCHvowGek2J", "port": 8604, "rocksdb_block_cache_size": 49426737}

[JeffreyDevloo]

Okay, the usage is based upon the asd their reported size and their reported usage The issue is that the size of the asd is not correct/possible?

[dejonghb]

There are 2 things:

• "capacity": 202451916800 verus 100.00 GiB size (in model)
• usage & total size exceed modelled 100.00 GiB as well

Guess df -B 1 is used on the mountpoint to find out about used / full size, not what asd/alba reports

(Note: running in lxd container and using directories on a disk, not a total device for itself)

[wimpers]

Currently LXD containers for ASDs are not supported. Something to tackle when we want to support this use case.