Error messages needs to be improved when invalid inputs are passed

[preethisn]

Error messages needs to be improved when invalid args are passed- having similar error messages will not help user to troubleshoot the issues with invalid inputs.

We expect error validation also to be taken care and throw messages accordingly which can help users to identify their mistake.

when host allow is not set and trying to connect on client side-
[root@ceph-ci-nvme-second-hgkhkr-node7 cephuser]# nvme connect -t tcp --traddr 10.0.209.223 -s 5001 -n nqn.2016-06.io.spdk:bdev0
Failed to write to /dev/nvme-fabrics: Input/output error
no controller found: failed to write to nvme-fabrics device

when invalidsubsystem nqn is passed-
[root@ceph-ci-nvme-second-hgkhkr-node7 cephuser]# nvme connect -t tcp --traddr 10.0.209.223 -s 5001 -n nqn.2016-06.io.spdk:bdev01
Failed to write to /dev/nvme-fabrics: Input/output error
no controller found: failed to write to nvme-fabrics device

when invalid gateway ip is passed-
[root@ceph-ci-nvme-second-hgkhkr-node7 cephuser]# nvme discover -t tcp -a 10.0.209.228 -s 5001
Failed to write to /dev/nvme-fabrics: Connection timed out
failed to add controller, error Unknown error -1

when invalid port number is passed
[root@ceph-ci-nvme-second-hgkhkr-node7 cephuser]# nvme discover -t tcp -a 10.0.209.223 -s 5000
Failed to write to /dev/nvme-fabrics: Connection timed out
failed to add controller, error Unknown error -1

[caroav]

@preethisn the examples that you give here, are from the linux nvme/tcp initiator. They are not something which is under any control of the GW AFAIK. So I'm not sure what is the requirement from the GW here?

[preethisn]

@preethisn the examples that you give here, are from the linux nvme/tcp initiator. They are not something which is under any control of the GW AFAIK. So I'm not sure what is the requirement from the GW here?

Yes, this is something on the client side improvements when nvme discover/connect commands are used with incorrect parameters. If its not feasable to fix these RFE's on client side we can document on the behaviour

[caroav]

@preethisn fixing it is not a part of this project. It is related to Linux initiator implementation. If you have no objections, I suggest to close this issue. I'm not sure if we need to keep it open for documentation, because I don't think that documenting this issues is also a part of this project.

[preethisn]

@preethisn fixing it is not a part of this project. It is related to Linux initiator implementation. If you have no objections, I suggest to close this issue. I'm not sure if we need to keep it open for documentation, because I don't think that documenting this issues is also a part of this project.

We are closing this bug as not applicable/invalid for fixing