Size of Zvol inc

[sandiphw]

We are using FreeNAS-11.1-U1. We have created multiple zvol of different sizes.

For eg, we have initially created a zvol of size 2T. But now it become 3.33T

pool1/data/smbdata/smb1 3.33T 2.05T 1.23T
pool1/data/smbdata/smb1@auto-20190315.2000-4w 1.37G - 1.22T
pool1/data/smbdata/smb1@auto-20190316.2000-4w 322M - 1.22T
pool1/data/smbdata/smb1@auto-20190317.2000-4w 327M - 1.22T
pool1/data/smbdata/smb1@auto-20190318.2000-4w 2.11G - 1.22T
pool1/data/smbdata/smb1@auto-20190319.2000-4w 1.69G - 1.22T
pool1/data/smbdata/smb1@auto-20190320.2000-4w 307M - 1.22T
pool1/data/smbdata/smb1@auto-20190321.2000-4w 413M - 1.22T
pool1/data/smbdata/smb1@auto-20190322.2000-4w 1012M - 1.23T
pool1/data/smbdata/smb1@auto-20190323.2000-4w 457M - 1.23T
pool1/data/smbdata/smb1@auto-20190324.2000-4w 399M - 1.23T
pool1/data/smbdata/smb1@auto-20190325.2000-4w 3.30G - 1.23T
pool1/data/smbdata/smb1@auto-20190326.2000-4w 2.84G - 1.23T
pool1/data/smbdata/smb1@auto-20190327.2000-4w 3.96G - 1.23T
pool1/data/smbdata/smb1@auto-20190328.2000-4w 3.73G - 1.23T
pool1/data/smbdata/smb1@auto-20190329.2000-4w 722M - 1.23T
pool1/data/smbdata/smb1@auto-20190330.2000-4w 296M - 1.23T
pool1/data/smbdata/smb1@auto-20190331.2000-4w 279M - 1.23T
pool1/data/smbdata/smb1@auto-20190401.2000-4w 9.57G - 1.23T
pool1/data/smbdata/smb1@auto-20190402.2000-4w 5.54G - 1.23T
pool1/data/smbdata/smb1@auto-20190403.2000-4w 1.98G - 1.23T
pool1/data/smbdata/smb1@auto-20190404.2000-4w 2.83G - 1.23T
pool1/data/smbdata/smb1@auto-20190405.2000-4w 2.76G - 1.23T
pool1/data/smbdata/smb1@auto-20190406.2000-4w 251M - 1.23T
pool1/data/smbdata/smb1@auto-20190407.2000-4w 258M - 1.23T
pool1/data/smbdata/smb1@auto-20190408.2000-4w 2.36G - 1.23T
pool1/data/smbdata/smb1@auto-20190409.2000-4w 3.09G - 1.23T
pool1/data/smbdata/smb1@auto-20190410.2000-4w 4.90G - 1.23T
pool1/data/smbdata/smb1@auto-20190411.2000-4w 747M - 1.23T

This way our 4 X 2T zvol now occupying more than 12 T.

Why the size in constantly increasing?

Regards,

Sandip

 

[garm]

What do you mean with zvol?

What you show is snapshots, they will increase indefinitely as long as the pool is written too and they remain.

 

[sandiphw]

I had create a zvol pool1/data/smbdata/smb1 of size 2 TB. Now its occupying 3.3 TB. Sizes of the all snapshots is less than 100GB. Then why this volume is occupying 3.3 TB?

 

[garm]

A zfs volume is a block device, a dataset is a filesystem in a pool. Datasets can have quotas set, volumes have a set size that can be provisioned or not. Which one do you have?

 

[sandiphw]

These are ZFS volumes. Please see the below image. Why smb1, smb2 and smb3 are using more than 3TB each when I have created zvol of 2TB. As a result we are running out of space. Before Freenas, we used Nexenta where we never faced this issue.

 

[garm]

What filesystem are you using on them?
What is the blocksize?
What is the filesystem cluster size?
You need to match them or zfs will allow the filesystem to effectively grow larger then the allocation.

Why bother with zvol instead of just use dataset and let ZFS worry about the filesystem bit?

 

[sandiphw]

We have shared these zvols through iSCSI and created ext3 filesystem of 4096 block size.

 

[garm]

And what is your zvol block size?

 

[garm]

Also, please post the print of zfs get all pool1/data/smbdata/smb1

 

[sandiphw]

Also, please post the print of zfs get all pool1/data/smbdata/smb1

NAME PROPERTY VALUE SOURCE
pool1/data/smbdata/smb1 type volume -
pool1/data/smbdata/smb1 creation Fri Feb 27 22:03 2015 -
pool1/data/smbdata/smb1 used 3.33T -
pool1/data/smbdata/smb1 available 2.12T -
pool1/data/smbdata/smb1 referenced 1.23T -
pool1/data/smbdata/smb1 compressratio 1.24x -
pool1/data/smbdata/smb1 reservation none default
pool1/data/smbdata/smb1 volsize 1.95T local
pool1/data/smbdata/smb1 volblocksize 16K -
pool1/data/smbdata/smb1 checksum on default
pool1/data/smbdata/smb1 compression lz4 inherited from pool1
pool1/data/smbdata/smb1 readonly off default
pool1/data/smbdata/smb1 copies 1 default
pool1/data/smbdata/smb1 refreservation 1.98T local
pool1/data/smbdata/smb1 primarycache all default
pool1/data/smbdata/smb1 secondarycache all default
pool1/data/smbdata/smb1 usedbysnapshots 123G -
pool1/data/smbdata/smb1 usedbydataset 1.23T -
pool1/data/smbdata/smb1 usedbychildren 0 -
pool1/data/smbdata/smb1 usedbyrefreservation 1.98T -
pool1/data/smbdata/smb1 logbias latency default
pool1/data/smbdata/smb1 dedup off default
pool1/data/smbdata/smb1 mlslabel -
pool1/data/smbdata/smb1 sync standard default
pool1/data/smbdata/smb1 refcompressratio 1.23x -
pool1/data/smbdata/smb1 written 3.41G -
pool1/data/smbdata/smb1 logicalused 1.65T -
pool1/data/smbdata/smb1 logicalreferenced 1.50T -

This iscsi volume is mounted in CentOS system. There this volume is showing as 2T only.