stardog data add throws filenotfound exception when cluster node cannot access file locally

[Jacobus Geluk]

When loading a file using the CLI without using the --server-side option, the cluster node that gets the request still assumes its a local file. It shows an error in the local stardog.log file (No such file or directory) that does not get propagated to the CLI which is confusing since finding the reason for the error then requires scanning all stardog.log files on all cluster nodes.

This is with 4.0-RC3 EE

[Evren Sirin]

Hi Jacobus,

Is it the case that you run this command on a node that has the file
but is not the coordinator?

Stardog enables --server-side automatically when you are adding data
to localhost and the coordinator will replicate the file contents to
other nodes. So this would have worked if the file exists only in the
coordinator and not the other nodes. This is still confusing behavior
and we'll make some adjustments there (issue #2661). The workaround is
to use something other than localhost in the connection URL, e.g.
using 127.0.0.1 will not cause server-side option to be enabled
automatically.

Best,
Evren

> --
> -- --
> You received this message because you are subscribed to the C&P "Stardog"
> group.
> To post to this group, send email to sta...@clarkparsia.com
> To unsubscribe from this group, send email to
> stardog+u...@clarkparsia.com
> For more options, visit this group at
> http://groups.google.com/a/clarkparsia.com/group/stardog?hl=en

[Jacobus Geluk]

Aha ok, thanks for that explanation, I fixed it earlier by using the HAProxy address but then the file actually gets transported over the network a few more times then necessary I guess.