This file is indexed.

/usr/share/doc/tgt/README.rbd is in tgt 1:1.0.51-1.

This file is owned by root:root, with mode 0o644.

The actual contents of the file can be viewed below.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
The 'rbd' backing-store driver provides access to Ceph (http://ceph.com)
'RADOS block devices (rbd)" through tgtd.

Ceph is a distributed storage system that provides clustered redundant
object storage; one of the ways to use such storage is through the
abstraction of a "block device".  rbd devices can be mapped through a
kernel driver and show up as block devices directly, or they can be used
through userland libraries to provide backing store for things like QEMU
virtual disks.

The bs_rbd backing-store driver allows tgtd to use rbd images as backing
storage without requiring the 'rbd' kernel block device driver, and so
avoids kernel version dependencies and the extra context switching between
kernel and userland.  It also allows for userland caching of block images,
managed by librbd.

To use bs_rbd, select --bstype rbd, and set --backing-store to an rbd
image specifier.  This is of the form

	[poolname/]imagename[@snapname]

where poolname is the RADOS pool containing the rbd image (defaults to
"rbd"), imagename is the image you want to export, and snapname is the
optional (readonly) snapshot of the image imagename.  (Obviously when
a snapshot is exported, writes are not supported.)

You may also want to use the --bsopts parameter to pass Ceph-specific
parameters to bs_rbd.  Accepted keywords are

conf=<path to ceph.conf file>

This sets the path to the ceph.conf file for the cluster that
stgt should connect to for this target/lun creation (luns may
use different cluster connections).

id=<client id>

This sets the id portion of the ceph "client name" (which is
"type.id").  Type is fixed at "client", and the default name is
"client.admin", but you probably want this to set id to
something like "tgt" so that the name of the ceph client is
"client.tgt".  This lets you control permissions differently
for the tgt client compared to others, and sets the default log path, etc.
See the Ceph documentation regarding client names.

cluster=<cluster name>

This sets the Ceph cluster name, if you have multiple clusters or
if your cluster name is anything other than "ceph".
This is in turn used by librados to find the conf file and key files.

To specify multiple options, separate them with ';', and since you are,
make sure to quote the option string to protect the semicolon from
the shell:

	--bsopts "conf=/usr/local/ceph/ceph.conf;id=tgt"