Ask Your Question

Revision history [back]

click to hide/show revision 1
initial version

LVM with another driver

I set up block service (Cinder) on Ubuntu machine by following the OpenStack guide for it (http://docs.openstack.org/icehouse/install-guide/install/apt/content/ch_cinder.html). As part of cinder-volume package installation, it asked to install lvm and set up an additional disk for its use.

apt-get install lvm2

pvcreate /dev/sdb

vgcreate cinder-volumes /dev/sdb

apt-get install cinder-volume

.

And then I installed another driver which went through fine. Cinder.conf file is populated correctly with volume backend name, enabled_backends, etc. I created a volume type and linked it to this driver via volume_backend_name. When I run volume create command by specifying volume_type id, cinder-api and cinder-volume logs shows that the volume has been created successfully and volume status is available. But cinder-scheduler log shows that ''the volume service is down or disabled. (host: HOSTNAME)''. Also there are no log lines in cinder-volume log indicating that it is contacting the backend array to create the volume and the volume is not created on the array as suspected.

cinder service-list +------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+ | Binary | Host | Zone | Status | State | Updated_at | +------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+

| cinder-scheduler | HOSTNAME | nova | enabled | up | 2014-08-07T07:17:06.000000 |

| cinder-volume | HOSTNAME | nova | enabled | down | 2014-08-07T06:54:47.000000 |

| cinder-volume | HOSTNAME@driver_iscsi | nova | enabled | up | 2014-08-07T07:17:01.000000 |

From above command, I observed that service is down. But I don’t understand why it is picking that cinder-volume service instead of selecting the driver one's. I manually disabled it so that scheduler might pick the second one. .

cinder service-list +------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+ | Binary | Host | Zone | Status | State | Updated_at | +------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+

| cinder-scheduler | HOSTNAME | nova | enabled | up | 2014-08-07T07:17:06.000000 |

| cinder-volume | HOSTNAME | nova | disabled | down | 2014-08-07T06:54:47.000000 |

| cinder-volume | HOSTNAME@driver_iscsi | nova | enabled | up | 2014-08-07T07:17:01.000000 |

But this time, I don’t see any error at all in scheduler log and the volume status is available. It does not look like it is using my driver to create volume even though I have specified the correct volume_type id. So, it is not selecting the proper volume service. .

My question is :

Do we actually see two cinder-volume services in this case ?

Why cinder-scheduler is not picking the correct one even when specified with volume_type id.?

Please help me resolve this!

.

cinder.conf file :

[DEFAULT]

rootwrap_config = /etc/cinder/rootwrap.conf api_paste_confg = /etc/cinder/api-paste.ini iscsi_helper = tgtadm volume_name_template = volume-%s volume_group = cinder-volumes verbose = True auth_strategy = keystone state_path = /var/lib/cinder lock_path = /var/lock/cinder volumes_dir = /var/lib/cinder/volumes rpc_backend = cinder.openstack.common.rpc.impl_kombu rabbit_host = controller rabbit_port = 5672 rabbit_userid = guest rabbit_password = password

enabled_backends = my_driver

[database]

connection = mysql://cinder:CINDER_DBPASS@controller/cinder

[keystone_authtoken]

auth_uri = http://controller:5000 auth_host = controller auth_port = 35357 auth_protocol = http admin_tenant_name = service admin_user = cinder admin_password = password

[my_driver]

volume_driver = cinder.volume.drivers.my_driver volume_backend_name = VOLUME_BACKEND_NAME OTHER LOGIN INFO NEEDED BY DRIVER

LVM with another driver

I set up block service (Cinder) on Ubuntu machine by following the OpenStack guide for it (http://docs.openstack.org/icehouse/install-guide/install/apt/content/ch_cinder.html). As part of cinder-volume package installation, it asked to install lvm and set up an additional disk for its use.

apt-get install lvm2

pvcreate /dev/sdb

vgcreate cinder-volumes /dev/sdb

apt-get install cinder-volume

.

And then I installed another driver which went through fine. Cinder.conf file is populated correctly with volume backend name, enabled_backends, etc. I created a volume type and linked it to this driver via volume_backend_name. When I run volume create command by specifying volume_type id, cinder-api and cinder-volume logs shows that the volume has been created successfully and volume status is available. But cinder-scheduler log shows that ''the volume service is down or disabled. (host: HOSTNAME)''. Also there are no log lines in cinder-volume log indicating that it is contacting the backend array to create the volume and the volume is not created on the array as suspected.

cinder service-list +------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+ | Binary | Host | Zone | Status | State | Updated_at | +------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+

| cinder-scheduler | HOSTNAME | nova | enabled | up | 2014-08-07T07:17:06.000000 |

| cinder-volume | HOSTNAME | nova | enabled | down | 2014-08-07T06:54:47.000000 |

| cinder-volume | HOSTNAME@driver_iscsi HOSTNAME@my_driver | nova | enabled | up | 2014-08-07T07:17:01.000000 |

From above command, I observed that service is down. But I don’t understand why it is picking that cinder-volume service instead of selecting the driver one's. I manually disabled it so that scheduler might pick the second one. .

cinder service-list +------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+ | Binary | Host | Zone | Status | State | Updated_at | +------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+

| cinder-scheduler | HOSTNAME | nova | enabled | up | 2014-08-07T07:17:06.000000 |

| cinder-volume | HOSTNAME | nova | disabled | down | 2014-08-07T06:54:47.000000 |

| cinder-volume | HOSTNAME@driver_iscsi HOSTNAME@my_driver | nova | enabled | up | 2014-08-07T07:17:01.000000 |

But this time, I don’t see any error at all in scheduler log and the volume status is available. It does not look like it is using my driver to create volume even though I have specified the correct volume_type id. So, it is not selecting the proper volume service. .

My question is :

Do we actually see two cinder-volume services in this case ?

Why cinder-scheduler is not picking the correct one even when specified with volume_type id.?

Please help me resolve this!

.

cinder.conf file :

[DEFAULT]

rootwrap_config = /etc/cinder/rootwrap.conf api_paste_confg = /etc/cinder/api-paste.ini iscsi_helper = tgtadm volume_name_template = volume-%s volume_group = cinder-volumes verbose = True auth_strategy = keystone state_path = /var/lib/cinder lock_path = /var/lock/cinder volumes_dir = /var/lib/cinder/volumes rpc_backend = cinder.openstack.common.rpc.impl_kombu rabbit_host = controller rabbit_port = 5672 rabbit_userid = guest rabbit_password = password

enabled_backends = my_driver

[database]

connection = mysql://cinder:CINDER_DBPASS@controller/cinder

[keystone_authtoken]

auth_uri = http://controller:5000 auth_host = controller auth_port = 35357 auth_protocol = http admin_tenant_name = service admin_user = cinder admin_password = password

[my_driver]

volume_driver = cinder.volume.drivers.my_driver volume_backend_name = VOLUME_BACKEND_NAME OTHER LOGIN INFO NEEDED BY DRIVER

LVM with another driver

I set up block service (Cinder) on Ubuntu machine by following the OpenStack guide for it (http://docs.openstack.org/icehouse/install-guide/install/apt/content/ch_cinder.html). As part of cinder-volume package installation, it asked to install lvm and set up an additional disk for its use.

apt-get install lvm2

pvcreate /dev/sdb

vgcreate cinder-volumes /dev/sdb

apt-get install cinder-volume

.

And then I installed another driver which went through fine. Cinder.conf file is populated correctly with volume backend name, enabled_backends, etc. I created a volume type and linked it to this driver via volume_backend_name. When I run volume create command by specifying volume_type id, cinder-api and cinder-volume logs shows that the volume has been created successfully and volume status is available. But cinder-scheduler log shows that ''the volume service is down or disabled. (host: HOSTNAME)''. Also there are no log lines in cinder-volume log indicating that it is contacting the backend array to create the volume and the volume is not created on the array as suspected.

cinder service-list +------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+ | Binary | Host | Zone | Status | State | Updated_at | +------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+

| cinder-scheduler | HOSTNAME | nova | enabled | up | 2014-08-07T07:17:06.000000 |

| cinder-volume | HOSTNAME | nova | enabled | down | 2014-08-07T06:54:47.000000 |

| cinder-volume | HOSTNAME@my_driver | nova | enabled | up | 2014-08-07T07:17:01.000000 |

From above command, I observed that service is down. But I don’t understand why it is picking that cinder-volume service instead of selecting the driver one's. I manually disabled it so that scheduler might pick the second one. .

cinder service-list +------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+ | Binary | Host | Zone | Status | State | Updated_at | +------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+

| cinder-scheduler | HOSTNAME | nova | enabled | up | 2014-08-07T07:17:06.000000 |

| cinder-volume | HOSTNAME | nova | disabled | down | 2014-08-07T06:54:47.000000 |

| cinder-volume | HOSTNAME@my_driver | nova | enabled | up | 2014-08-07T07:17:01.000000 |

But this time, I don’t see any error at all in scheduler log and the volume status is available. It does not look like it is using my driver to create volume even though I have specified the correct volume_type id. So, it is not selecting the proper volume service. .

My question is :

Do we actually see two cinder-volume services in this case ?

Why cinder-scheduler is not picking the correct one even when specified with volume_type id.?

Please help me resolve this!

.

cinder.conf file :

[DEFAULT]

rootwrap_config = /etc/cinder/rootwrap.conf api_paste_confg = /etc/cinder/api-paste.ini iscsi_helper = tgtadm volume_name_template = volume-%s volume_group = cinder-volumes verbose = True auth_strategy = keystone state_path = /var/lib/cinder lock_path = /var/lock/cinder volumes_dir = /var/lib/cinder/volumes rpc_backend = cinder.openstack.common.rpc.impl_kombu rabbit_host = controller rabbit_port = 5672 rabbit_userid = guest rabbit_password = password

enabled_backends = my_driver

[database]

connection = mysql://cinder:CINDER_DBPASS@controller/cinder

[keystone_authtoken]

auth_uri = http://controller:5000 auth_host = controller auth_port = 35357 auth_protocol = http admin_tenant_name = service admin_user = cinder admin_password = password

[my_driver]

volume_driver = cinder.volume.drivers.my_driver volume_backend_name = VOLUME_BACKEND_NAME OTHER LOGIN CONNECTION INFO NEEDED BY DRIVER

click to hide/show revision 4
No.4 Revision

LVM with another driver

I set up block service (Cinder) on Ubuntu machine by following the OpenStack guide for it (http://docs.openstack.org/icehouse/install-guide/install/apt/content/ch_cinder.html). As part of cinder-volume package installation, it asked to install lvm and set up an additional disk for its use.

apt-get install lvm2

lvm2 pvcreate /dev/sdb

/dev/sdb vgcreate cinder-volumes /dev/sdb

/dev/sdb apt-get install cinder-volume

cinder-volume

.

And then I installed another driver which went through fine. Cinder.conf file is populated correctly with volume backend name, enabled_backends, etc. I created a volume type and linked it to this driver via volume_backend_name. When I run volume create command by specifying volume_type id, cinder-api and cinder-volume logs shows that the volume has been created successfully and volume status is available. But cinder-scheduler log shows that ''the volume service is down or disabled. (host: HOSTNAME)''. Also there are no log lines in cinder-volume log indicating that it is contacting the backend array to create the volume and the volume is not created on the array as suspected.

cinder service-list
+------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+
|      Binary      |                        Host                              | Zone |  Status | State |         Updated_at         |
+------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+

+------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+ | cinder-scheduler | HOSTNAME | nova | enabled | up | 2014-08-07T07:17:06.000000 |

| | cinder-volume | HOSTNAME | nova | enabled | down | 2014-08-07T06:54:47.000000 |

| | cinder-volume | HOSTNAME@my_driver | nova | enabled | up | 2014-08-07T07:17:01.000000 |

|

From above command, I observed that that service is down. But I don’t don’t understand why it is picking that that cinder-volume service instead of of selecting the driver one's. I manually manually disabled it so that scheduler might might pick the second one. .one.

cinder service-list
+------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+
|      Binary      |                        Host                              | Zone |  Status | State |         Updated_at         |
+------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+

+------------------+-----------------------------------------------------------------+------+---------+-------+------------------------+ | cinder-scheduler | HOSTNAME | nova | enabled | up | 2014-08-07T07:17:06.000000 |

| | cinder-volume | HOSTNAME | nova | disabled | down | 2014-08-07T06:54:47.000000 |

| | cinder-volume | HOSTNAME@my_driver | nova | enabled | up | 2014-08-07T07:17:01.000000 |

|

But this time, I don’t see any error error at all in scheduler log and the volume volume status is available. It does not look look like it is using my driver to create create volume even though I have specified specified the correct volume_type id. So, it is is not selecting the proper volume service. .volume service.

My question is :

Do we actually see two cinder-volume cinder-volume services in this case ?

Why cinder-scheduler is not picking picking the correct one even when specified specified with volume_type id.?

Please help me resolve this!

.

cinder.conf file :

[DEFAULT]

[DEFAULT]

rootwrap_config = /etc/cinder/rootwrap.conf
api_paste_confg = /etc/cinder/api-paste.ini
iscsi_helper = tgtadm
volume_name_template = volume-%s
volume_group = cinder-volumes
verbose = True
auth_strategy = keystone
state_path = /var/lib/cinder
lock_path = /var/lock/cinder
volumes_dir = /var/lib/cinder/volumes
rpc_backend = cinder.openstack.common.rpc.impl_kombu
rabbit_host = controller
rabbit_port = 5672
rabbit_userid = guest
rabbit_password = password

password enabled_backends = my_driver

[database]

my_driver [database] connection = mysql://cinder:CINDER_DBPASS@controller/cinder

[keystone_authtoken]

mysql://cinder:CINDER_DBPASS@controller/cinder [keystone_authtoken] auth_uri = http://controller:5000 auth_host = controller auth_port = 35357 auth_protocol = http admin_tenant_name = service admin_user = cinder admin_password = password

[my_driver]

password [my_driver] volume_driver = cinder.volume.drivers.my_driver volume_backend_name = VOLUME_BACKEND_NAME OTHER CONNECTION INFO NEEDED BY DRIVER

DRIVER