ceph/ceph-iscsi-tools

Name: ceph-iscsi-tools

Owner: Ceph

Description: Useful tools for a ceph/iscsi gateway environment

Created: 2016-09-29 02:53:42.0

Updated: 2018-01-22 18:47:42.0

Pushed: 2017-11-17 02:20:58.0

Homepage: null

Size: 270

Language: Python

GitHub Committers

User	Most Recent Commit	# Commits

Other Committers

User	Email	Most Recent Commit	# Commits

README

ceph-iscsi-tools

This repo provides some helper tools for ceph/iscsi environments.

gwtop

This is a top-like tool intended to be installed on each gateway node. It provides an aggregated view of I/O from each
gateway node, combined with LIO information - matching the host device to a client using the device over iSCSI.
Performance metrics are sourced from performance co-pilot (pmcd), so this service needs to be running on each gateway
node.

The output varies dependent on the pmda providing the metrics.

For krbd based devices;

p  2/2 Gateways   CPU% MIN:  4 MAX:  5    Network Total In:    2M  Out:    3M   10:20:09
city:   8G    Disks:   8   IOPS:  500   Clients:  1   Ceph: HEALTH_OK          OSDs:   3
.Image     Src  Device   Size     r/s     w/s    rMB/s     wMB/s    await  r_await  w_await  Client
i.t1703          rbd0    500M       0       0     0.00      0.00     0.00     0.00     0.00              
i.testme1        rbd5    500M       0       0     0.00      0.00     0.00     0.00     0.00                  
i.testme2        rbd2    500M       0       0     0.00      0.00     0.00     0.00     0.00                     
i.testme3        rbd3    500M       0       0     0.00      0.00     0.00     0.00     0.00              
i.testme5        rbd1    500M       0       0     0.00      0.00     0.00     0.00     0.00              
myhost_1    T    rbd4      4G     500       0     1.95      0.00     2.37     2.37     0.00  rh460p(CON)         
test_2           rbd6      1G       0       0     0.00      0.00     0.00     0.00     0.00              
testme           rbd7    500M       0       0     0.00      0.00     0.00     0.00     0.00              

For user backed storage (TCMU devices)

p  2/2 Gateways   CPU% MIN:  4 MAX:  5    Network Total In:    2M  Out:    3M   10:20:00
city:   8G    Disks:   8   IOPS:  503   Clients:  1   Ceph: HEALTH_OK          OSDs:   3
.Image       Src    Size     iops     rMB/s     wMB/s   Client
i.t1703             500M        0      0.00      0.00                       
i.testme1           500M        0      0.00      0.00                       
i.testme2           500M        0      0.00      0.00                       
i.testme3           500M        0      0.00      0.00                       
i.testme5           500M        0      0.00      0.00                       
myhost_1      T       4G      504      1.95      0.00   rh460p(CON)         
test_2                1G        0      0.00      0.00                       
testme              500M        0      0.00      0.00                       

gwtop determines the gateway configuration by looking at the following sources (in order)

• the invocation used the -g host-1,host-2 over-ride
  
• the user has a .gwtop.rc file in the root of their home directory
  
• the environment was created by the ceph-iscsi-ansible project, and has committed state to the rbd/gateway.conf object
  

An example of the 'rc' file is provided in the /usr/share/doc directory.

Invocation provides the following options;

t@ceph-1 ~]# gwtop -h
e: gwtop [-h] [-b] [-c CONFIG_OBJECT] [-g GATEWAYS]
         [-i {1,2,3,4,5,6,7,8,9}] [-d] [-m {text}] [-p {dm,lio}]
         [-s {image,rbd_name,reads,writes,await,io_source}] [-r] [-v]

 iSCSI gateway performance metrics

onal arguments:
, --help            show this help message and exit
, --busy-only       show only active devices (iops > 0)
 CONFIG_OBJECT, --config-object CONFIG_OBJECT
                    pool and object name holding the gateway config object
                    (pool/object_name)
 GATEWAYS, --gateways GATEWAYS
                    comma separated iscsi gateway server names
 {1,2,3,4,5,6,7,8,9}, --interval {1,2,3,4,5,6,7,8,9}
                    monitoring interval (secs)
, --debug           run with additional debug
 {text}, --mode {text}
                    output mode
 {dm,lio}, --provider {dm,lio}
                    pcp provider type lio or dm
 {image,rbd_name,reads,writes,await,io_source}, --sortkey {image,rbd_name,reads,writes,await,io_source}
                    sort key sequence
, --reverse         use reverse sort when displaying the stats
, --version         show program's version number and exit