Frequently Asked Questions

Q. Where can I find the GPL sources that were used for the Add-ons Ext2/XFS?

It is hosted in

Q. Does Crossmeta support EXT3 file system?

Yes. Crossmeta EXT2fs driver supports ext3 file system and is treated as ext2 without journaling.This would work only if the EXT3 file system was cleanly shutdown and has no pending journal file to be applied.
You would see the following message in dmesg output

EXT2-fs warning (device da1): ext2_read_super: mounting ext3 filesystem as ext2

If there is active journal file then ext2 would refuse to mount the file system as it lacks journaling capability.

EXT2-fs: da1: couldn't mount because of unsupported optional features (4).

In that case e2fsck is required to replay journal file and then mount with Crossmeta ext2fs driver.

Q. Is Ext2 filesystem with dynamic inode size of 256 bytes supported?

Yes updated driver to expanded inode size is available for testing and will be posted in Files section.

Q. How to access Linux file systems under LVM2 volumes?

There is no automatic scanning and recognition of LVM2 volumes in Crossmeta. However one can manually define a loop partition device based on LVM volume table information.

To construct a loop device for the lvm volume, obtain the start and size of the logical volume available from Linux as follows:

  1. In Linux get a listing of LVM volumes
    # dmsetup ls
    lvmtest-winfs2    (253, 13)
    lvmtest-winfs1    (253, 12)

To obtain the table of lvmtest-winfs2

# dmsetup table lvmtest-winfs2
0 163840 linear 7:1 206848

It shows start is at 206848 sectors, up to a size of 163840 sectors. For Crossmeta losetup command these have to be converted to bytes.

Based on this information make loop device in Crossmeta as follows:

If the disk image is in file C:\TEMP\lvmtest.img

v:\bin> losetup -o 105906176 -s 83886080 /dev/loop1 C:\TEMP\lvmtest.img

If the disk image is hard drive recognized in Windows as \\.\\Harddisk1, then

v:\bin> losetup -o 105906176 -s 83886080 /dev/loop1 Harddisk1

This will create next available device loop1
To mount this file system

# mount -t ext2fs /dev/loop1 /mnt

Q. What to do if NFS is not being used?

Normally NFS file system driver cxnfs.sys is started only if NFS shares are specified in /etc/exports. If NFS service is not required you can manually stop cxnfs service as follows:

c:\ net stop cxnfs

Q.How to access the NFS share of Crossmeta server from Linux or other NFS clients?

Since Crossmeta NFS supports only NFS v2 or v3 in UDP transport, the following should work.

# mount -t nfs -o vers=3,proto=udp crossmeta:/mnt  /nfstest