periodically check space used with du

[aleitner]

No description provided.

[littleskunk]

Please check the free space as well. The allocated size is only validated once on start.

[braydonf]

LGTM, have done a bit of testing. Let's test some more before we make a release though.

[ne0ark]

This seems to be crashing daemon start. Is there a way to roll back to old core version? There is nothing logged. It seems like if du takes a long daemon probably times out waiting for the node to start.

[braydonf]

This only adds the methods to be used, and doesn't run them. What version are you running of daemon? Pulling from master branch or from a tagged release?

[ne0ark]

I just upgrade to:

storjshare --version
daemon: 4.0.1, core: 7.0.2, protocol: 1.2.0
/dev/sda3        17T  7.9T  7.6T  51% /

At 8 TB two nodes I get:

failed to start share, reason: invalid storage size

This started happening after 4.0.0 Daemon version. Currently, two node has ~7.4 TB used.

Reducing the node size to 7TB. Crashed the daemon and nothing gets logged (daemon.log @ debug level) accept:

{"level":"info","message":"attempting to start share with config at path /root/.config/storjshare/configs/<id>.json","timestamp":"2017-09-07T13:35:20.816Z"}

[braydonf]

That may be another issue.

[braydonf]

@ne0ark There was an issue with 4.0.0 and fixed in 4.0.1 of storjshare-daemon, maybe the node modules need to be updated?

[ne0ark]

nodejs --version
v6.11.3

I am running kfs -d Storj/Storj01/sharddata.kfs compact to rule out corruption. But something else might be at play here. All modules should have been auto upgraded. I used npm install --global storjshare-daemon to install.

[ne0ark]

@braydonf storj-archived/storjshare-daemon#247 Error log