Another replication question

[hescominsoon]

I am having a problem getting a remote freeans seeing the dataset i have setup for it to replicate to.
The remote dataset is Data/wthu files
The local dataset is wthu/wthu files

When i fill out the replication screen on the remote freenas I am using the following paramters:
wthu/wthu files
remote: Data/wthu files


Everything else is working fine except something with the dataset setup for replication because the error i get is:
Fialed: cannot open 'Data/wthu files/wthu files': dataset does not exist.

I am curious why it is calling out the extra wthu files...

 

[dlavigne]

Were you able to figure this out?

 

[Nick2253]

Can you post a screenshot of your replication settings?

 

[PhilipS]

The remote dataset that is created/written with the value entered in Remote ZFS Volume/Dataset with the local dataset name appended to it minus the volume name the local dataset is on.

So a remote dataset called tank/backup
and a local dataset called tank/mydata/important
will create/write to a remote dataset of tank/backup/mydata/important

Not sure why you're getting an error though, if the remote dataset does not exist, the replication script is supposed to create it.

 

[hescominsoon]

The remote dataset that is created/written with the value entered in Remote ZFS Volume/Dataset with the local dataset name appended to it minus the volume name the local dataset is on.

So a remote dataset called tank/backup
and a local dataset called tank/mydata/important
will create/write to a remote dataset of tank/backup/mydata/important

Not sure why you're getting an error though, if the remote dataset does not exist, the replication script is supposed to create it.

I finally just created the dataset it was calling out. here's a screenshot of my final working settings on the remote FreeNAS:

 

[Nick2253]

Wow, Windows XP. Now that's a flashback ;)

I'm wondering if it has something to do with the space in your dataset name. That might be breaking something in the dataset creation scripts used by the replication. Would you be willing to file a bug report on this?

 

[PhilipS]

The Remote ZFS Volume/Dataset setting should be just Data if you want the remote to be in Data/wthu files. As configured the remote will be Data/wthu files/wthu files

 

[hescominsoon]

The Remote ZFS Volume/Dataset setting should be just Data if you want the remote to be in Data/wthu files. As configured the remote will be Data/wthu files/wthu files

The way the documentation puts it and the way it is presented in the screen that's a bit counter-intuitive.

 

[hescominsoon]

Wow, Windows XP. Now that's a flashback ;)

I'm wondering if it has something to do with the space in your dataset name. That might be breaking something in the dataset creation scripts used by the replication. Would you be willing to file a bug report on this?

yeah this machine not only generates the webstream but also interfaces with some really old equipment that I cannot replicate the functionality until we replace the main mixing board in the main studio. We are planning to replace the main board soon...then that xp machine gets replaced as well..:)

 

[hescominsoon]

The Remote ZFS Volume/Dataset setting should be just Data if you want the remote to be in Data/wthu files. As configured the remote will be Data/wthu files/wthu files

The error i got though is the dataset didn't exist. The system refused to auto-create the dataset..i had to goto my local Freenas and manually create the dataset it was calling out. From what I have read it should have created the destination dataset but it didn't. I had to goto my local freenas and create the dataset tree it wanted.

 

[PhilipS]

The way the documentation puts it and the way it is presented in the screen that's more than a bit counter-intuitive.

Agreed, and because of this design, your dataset on the remote can't have a different name than the source and if nested, then it will be nested on the target. If this bothers you, you may want to put in a bug report about it.

 

[PhilipS]

The error i got though is the dataset didn't exist. The system refused to auto-create the dataset..i had to goto my local Freenas and manually create the dataset it was calling out. From what I have read it should have created the destination dataset but it didn't. I had to goto my local freenas and create the dataset tree it wanted.

Yeah, as Nick said, that may be a bug due to the space in the dataset name.

 

[PhilipS]

Do you have any replication related error messages in /var/log/debug.log?

(I think the error would be: too many arguments)

I think that the autorepl.py script is not handling spaces in dataset creation (There are no quotes being placed around the dataset name). It also doesn't look like it would see the new dataset on future replications (it won't put in remote_zfslist correctly) and try to create it again.

 

[hescominsoon]

here's the bug report:
https://redmine.ixsystems.com/issues/28657