Newbie - How to enable compression? Lost...

[tine1999]

Hi,
I am a newbie at all this and lost in the dataset thingy. I have set-up a zfs raided pool with 5 disks called tank1. I have been using it for a while as a simple CIFS share.
Now, that the pool is starting to fill-in, I just realize that I should have enabled compression.... upsi.....

Anyhow, since I still have some room left in the pool, I was thinking I could perhaps:

1 - Create a dataset in the pool tank1 (let's call it T1Dataset with no quota)
2 - Enable compression in the dataset T1Dataset - gzip 6
3 - Move all the files from tank1 into T1Dataset

The gist of it is that as files are "removed" from tank1 and compressed into T1Dataset, as space frees from tank1, this space can be used by T1Dataset. Since the actual space used by T1Dataset will be less than the uncompressed files in tank1, I am guessing it should work.

However, I am lost at the whole datast thingy.
To create a dataset in GUI, I go to Disk / ZFS / Datasets and create one with compression enabled. Save.
So far, so good. And now what?

If I check Disk / ZFS / Dataset / Information, is not listed there (only tank1)
If I check Disk / ZFS / Dataste is listed there
If I check Disk / ZFS / Configuration / Current, it is there
If I check Disk / ZFS / Configuration / Detected, it's not there

Furthermore, where is T1Dataset mounted? I can't find it anywhere.

What am I missing? Do I need to reboot or something for the dataset to be created?

Totally lost. Any help would be very appreciated! Tx!

[lux]

would not work

i spend some time to test zfs compression some month ago

there's no noticable higer compression for media files such as video/audio/pics

don't know your files but if any (compressed) media in there is would say it would not work

[johl]

Hi, you can see a dataset as an map and as a filesystem, in which you can set compression, permissions and split your snapshots. Hence, every dataset can create/be used as an snapshot (not the whole pool). If needed, run snapshot with -R to handle to whole pool for snapshot. Works perfect when you'll need to set quotas for different things for you needs. I use it for my time capsule "slice", my snapshots, and some dropbox-alike applications SparkleShare.

If you start and console or and ssh session, you can type:

Code: Select all

# zfs list -o space

Code: Select all

# zfs list -t snapshot

To se some of your usage in ifs.

Cheers Jonas

Hi,
I am a newbie at all this and lost in the dataset thingy. I have set-up a zfs raided pool with 5 disks called tank1. I have been using it for a while as a simple CIFS share.
Now, that the pool is starting to fill-in, I just realize that I should have enabled compression.... upsi.....

Anyhow, since I still have some room left in the pool, I was thinking I could perhaps:

1 - Create a dataset in the pool tank1 (let's call it T1Dataset with no quota)
2 - Enable compression in the dataset T1Dataset - gzip 6
3 - Move all the files from tank1 into T1Dataset

The gist of it is that as files are "removed" from tank1 and compressed into T1Dataset, as space frees from tank1, this space can be used by T1Dataset. Since the actual space used by T1Dataset will be less than the uncompressed files in tank1, I am guessing it should work.

However, I am lost at the whole datast thingy.
To create a dataset in GUI, I go to Disk / ZFS / Datasets and create one with compression enabled. Save.
So far, so good. And now what?

If I check Disk / ZFS / Dataset / Information, is not listed there (only tank1)
If I check Disk / ZFS / Dataste is listed there
If I check Disk / ZFS / Configuration / Current, it is there
If I check Disk / ZFS / Configuration / Detected, it's not there

Furthermore, where is T1Dataset mounted? I can't find it anywhere.

What am I missing? Do I need to reboot or something for the dataset to be created?

Totally lost. Any help would be very appreciated! Tx!

[aaronb]

You Datasets are Mounted within the pool they are created from.

When you created your ZFS file system, you created the RAIDz1, the Pool, and the Dataset.

The mount point by default is /mnt/Pool/Dataset
- in other words, the dataset looks like a folder within the Pool.
If you add another dataset to the pool, i.e. - newDataset, it also is mounted by default in the pool at /mnt/Pool/newDataSet

You should be able to share out the Pool and see both.
You should also be able to just share out the other dataset to see them both from the network.