)]}'
{"doc/source/deploy/user-guide.rst":[{"author":{"_account_id":6618,"name":"Ruby Loo","email":"opensrloo@gmail.com","username":"rloo"},"change_message_id":"1c444e1f1a6e1ce94dcfb1441745fbca291f1926","unresolved":false,"context_lines":[{"line_number":157,"context_line":".. figure:: ../images/deployment_steps.png\r"},{"line_number":158,"context_line":"   :alt: Deployment Steps\r"},{"line_number":159,"context_line":"|\r"},{"line_number":160,"context_line":"     These pre-requisites must be met before the deployment process:\r"},{"line_number":161,"context_line":"|\r"},{"line_number":162,"context_line":"|\r"},{"line_number":163,"context_line":"     a. Dependent packages to be configured on the compute node like dnsmasq, tftp-server, ipmi, syslinux etc for bare metal provisioning.\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"1ae5cdf2_9bc09e7d","line":160,"updated":"2014-06-20 02:54:54.000000000","message":"Starting from here, the pre-requisites seems to stop the flow, since the deploy process below is associated with the diagram above. How about moving this part up, so that it is the first paragraph (or the only part) of \"Understanding Bare Metal Deployment\".\n\nThen under the \u0027Deploy Process\u0027 section below, it can start with \u0027What happens when a boot instance...\u0027, followed by the diagram, followed by the numbered steps.\n\nWould that be better?","commit_id":"8a556c8641ce402518c6d7ff87a6e69cc20ac507"},{"author":{"_account_id":6618,"name":"Ruby Loo","email":"opensrloo@gmail.com","username":"rloo"},"change_message_id":"1c444e1f1a6e1ce94dcfb1441745fbca291f1926","unresolved":false,"context_lines":[{"line_number":176,"context_line":"Deploy Process\r"},{"line_number":177,"context_line":"-----------------\r"},{"line_number":178,"context_line":"|\r"},{"line_number":179,"context_line":"     1. Nova boot instance request comes in, messages the API which goes through the message queue\r"},{"line_number":180,"context_line":"|\r"},{"line_number":181,"context_line":"|\r"},{"line_number":182,"context_line":"     2. Nova scheduler applies filter and finds the eligible compute node.\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"1ae5cdf2_7b7d921b","line":179,"updated":"2014-06-20 02:54:54.000000000","message":"How about \u0027A boot instance request comes in via the Nova API, through the message queue, to the Nova scheduler.\u0027\n\nHmm, I don\u0027t know what others think, but I am wondering whether the message queue clutters up the diagram; I\u0027m not sure it adds much value.","commit_id":"8a556c8641ce402518c6d7ff87a6e69cc20ac507"},{"author":{"_account_id":6618,"name":"Ruby Loo","email":"opensrloo@gmail.com","username":"rloo"},"change_message_id":"1c444e1f1a6e1ce94dcfb1441745fbca291f1926","unresolved":false,"context_lines":[{"line_number":188,"context_line":"     4. Information about the bare metal node is retrieved from the bare metal database and the node is reserved.\r"},{"line_number":189,"context_line":"|\r"},{"line_number":190,"context_line":"|\r"},{"line_number":191,"context_line":"     5. Images from Glance are pulled down to the local disks.\r"},{"line_number":192,"context_line":"|\r"},{"line_number":193,"context_line":"|\r"},{"line_number":194,"context_line":"     6. Virtual interfaces are plugged in.\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"1ae5cdf2_1b09ee94","line":191,"updated":"2014-06-20 02:54:54.000000000","message":"disks -\u003e local disk of the Ironic conductor servicing the bare metal node.","commit_id":"8a556c8641ce402518c6d7ff87a6e69cc20ac507"},{"author":{"_account_id":6618,"name":"Ruby Loo","email":"opensrloo@gmail.com","username":"rloo"},"change_message_id":"1c444e1f1a6e1ce94dcfb1441745fbca291f1926","unresolved":false,"context_lines":[{"line_number":194,"context_line":"     6. Virtual interfaces are plugged in.\r"},{"line_number":195,"context_line":"|\r"},{"line_number":196,"context_line":"|\r"},{"line_number":197,"context_line":"     7. The bare metal driver deploys the bare metal node which is passed to the Ironic API and the conductor.\r"},{"line_number":198,"context_line":"|\r"},{"line_number":199,"context_line":"|\r"},{"line_number":200,"context_line":"     8. PXE driver prepares tftp bootloader.\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"1ae5cdf2_1b300e55","line":197,"updated":"2014-06-20 02:54:54.000000000","message":"How about \u0027Nova\u0027s ironic driver issues a deploy request via the Ironic API to the Ironic conductor servicing the bare metal node.\u0027","commit_id":"8a556c8641ce402518c6d7ff87a6e69cc20ac507"},{"author":{"_account_id":6618,"name":"Ruby Loo","email":"opensrloo@gmail.com","username":"rloo"},"change_message_id":"1c444e1f1a6e1ce94dcfb1441745fbca291f1926","unresolved":false,"context_lines":[{"line_number":200,"context_line":"     8. PXE driver prepares tftp bootloader.\r"},{"line_number":201,"context_line":"|\r"},{"line_number":202,"context_line":"|\r"},{"line_number":203,"context_line":"     9. The driver issues IPMI command to enable network boot of a node and power it on.\r"},{"line_number":204,"context_line":"|\r"},{"line_number":205,"context_line":"|\r"},{"line_number":206,"context_line":"     10. The DHCP boots the deploy ramdisk. The PXE driver actually copies the image over iSCSI to the physical node. It connects to the iSCSI end point, partitions volume, \"dd\" the image and closes the iSCSI connection. The deployment is done.\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"1ae5cdf2_7b275212","line":203,"updated":"2014-06-20 02:54:54.000000000","message":"There are too many drivers! :-)\n\nThis one is \u0027The IPMI driver\u0027.","commit_id":"8a556c8641ce402518c6d7ff87a6e69cc20ac507"},{"author":{"_account_id":8125,"name":"Michael Davies","email":"michael@the-davies.net","username":"mrda"},"change_message_id":"24a5cee142ac0d23abff64843038f577933d5e6d","unresolved":false,"context_lines":[{"line_number":141,"context_line":"     1.  The controller runs the identity service, management service, dashboard and the management portion of compute. It also contains associated API services, MySQL databases and messaging system.\r"},{"line_number":142,"context_line":"|\r"},{"line_number":143,"context_line":"|\r"},{"line_number":144,"context_line":"     2.  The Ironic RESTful api service is used to enroll hardware with attributes like MAC addresses, IPMI credentials etc. A cloud administrator usually enrolls this information for Ironic to manage the specific hardware.\r"},{"line_number":145,"context_line":"|\r"},{"line_number":146,"context_line":"|\r"},{"line_number":147,"context_line":"     3.  The compute node runs the Nova compute service, networking plug-in agent and Ironic conductor service. The Ironic conductor service does the bulk of the work. There can be multiple instances of the conductor service to support various class of drivers and also to manage fail over. Ideally, instances of conductor service should be on separate nodes. Each conductor can itself run many drivers to operate heterogeneous hardware. This is depicted in figure 1.3.2. The API exposes a list of supported drivers and the names of conductor hosts servicing them.\r"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_2afc5f0c","line":144,"updated":"2014-06-23 08:11:32.000000000","message":"Small nit: api is normally written API as it\u0027s an acronym.  It\u0027s listed as API elsewhere in this document.","commit_id":"cea73de582784fad8245e64c9cd8f1edbaafac46"},{"author":{"_account_id":8106,"name":"Haomeng,Wang","email":"wanghaomeng@gmail.com","username":"whaom"},"change_message_id":"431155b217c9acc1714915d90769ad613b71763f","unresolved":false,"context_lines":[{"line_number":157,"context_line":"     These pre-requisites must be met before the deployment process:\r"},{"line_number":158,"context_line":"|\r"},{"line_number":159,"context_line":"|\r"},{"line_number":160,"context_line":"     a. Dependent packages to be configured on the compute node like dnsmasq, tftp-server, ipmi, syslinux etc for bare metal provisioning.\r"},{"line_number":161,"context_line":"|\r"},{"line_number":162,"context_line":"|\r"},{"line_number":163,"context_line":"     b. Flavors to be created for the available hardware. Nova must know the flavor to boot from.\r"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_64fb007c","line":160,"updated":"2014-06-23 07:22:32.000000000","message":"dnsmasq is controlled by neutron already, so from ironic view, we dont depend it directly I think.","commit_id":"cea73de582784fad8245e64c9cd8f1edbaafac46"},{"author":{"_account_id":8106,"name":"Haomeng,Wang","email":"wanghaomeng@gmail.com","username":"whaom"},"change_message_id":"431155b217c9acc1714915d90769ad613b71763f","unresolved":false,"context_lines":[{"line_number":163,"context_line":"     b. Flavors to be created for the available hardware. Nova must know the flavor to boot from.\r"},{"line_number":164,"context_line":"|\r"},{"line_number":165,"context_line":"|\r"},{"line_number":166,"context_line":"     c. Images to be made available in Glance. Different images like deploy kernel and ramdisk images as well as machine images are required for bare metal deployment.\r"},{"line_number":167,"context_line":"|\r"},{"line_number":168,"context_line":"|\r"},{"line_number":169,"context_line":"     d. Hardware to be enrolled via Ironic RESTful API service.\r"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_64322060","line":166,"updated":"2014-06-23 07:22:32.000000000","message":"Can we list the image types which are required for Ironic here?\n\nbm-deploy-kernel/bm-deploy-ramdisk\n\nuser-image/user-image-vmlinuxz/user-image-initrd","commit_id":"cea73de582784fad8245e64c9cd8f1edbaafac46"},{"author":{"_account_id":8106,"name":"Haomeng,Wang","email":"wanghaomeng@gmail.com","username":"whaom"},"change_message_id":"431155b217c9acc1714915d90769ad613b71763f","unresolved":false,"context_lines":[{"line_number":169,"context_line":"     d. Hardware to be enrolled via Ironic RESTful API service.\r"},{"line_number":170,"context_line":"|\r"},{"line_number":171,"context_line":".. figure:: ../images/deployment_steps.png\r"},{"line_number":172,"context_line":"   :alt: Deployment Steps\r"},{"line_number":173,"context_line":"|\r"},{"line_number":174,"context_line":"Deploy Process\r"},{"line_number":175,"context_line":"-----------------\r"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_245bb8fa","line":172,"updated":"2014-06-23 07:22:32.000000000","message":"This is clear diagram, just a few comments about the diagram\n\n1. For step 7 or 8, there should be a poweron operation to start the physical server I think, which will use IPMI driver for power management.\n2. For the 11st step, it is reboot action, need depends on IPMI I think.","commit_id":"cea73de582784fad8245e64c9cd8f1edbaafac46"},{"author":{"_account_id":8106,"name":"Haomeng,Wang","email":"wanghaomeng@gmail.com","username":"whaom"},"change_message_id":"431155b217c9acc1714915d90769ad613b71763f","unresolved":false,"context_lines":[{"line_number":178,"context_line":"|\r"},{"line_number":179,"context_line":"|\r"},{"line_number":180,"context_line":"     2. Nova scheduler applies filter and finds the eligible compute node.\r"},{"line_number":181,"context_line":"|\r"},{"line_number":182,"context_line":"|\r"},{"line_number":183,"context_line":"     3. A spawn task is placed by the driver which contains all information such as which image to boot from etc. It invokes the driver.spawn from the virt layer of Nova compute.\r"},{"line_number":184,"context_line":"|\r"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_e44550b5","line":181,"updated":"2014-06-23 07:22:32.000000000","message":"How do you think if we need more details about the flavor\u0027s extra_specs data is required for bare metal, such as \u0027cpu_arch\u0027, \u0027baremetal:deploy_kernel_id\u0027 / \u0027baremetal:deploy_ramdisk_id\u0027 to match the target physical nodes by nova scheduler?","commit_id":"cea73de582784fad8245e64c9cd8f1edbaafac46"},{"author":{"_account_id":8125,"name":"Michael Davies","email":"michael@the-davies.net","username":"mrda"},"change_message_id":"24a5cee142ac0d23abff64843038f577933d5e6d","unresolved":false,"context_lines":[{"line_number":183,"context_line":"     3. A spawn task is placed by the driver which contains all information such as which image to boot from etc. It invokes the driver.spawn from the virt layer of Nova compute.\r"},{"line_number":184,"context_line":"|\r"},{"line_number":185,"context_line":"|\r"},{"line_number":186,"context_line":"     4. Information about the bare metal node is retrieved from the bare metal database and the node is reserved.\r"},{"line_number":187,"context_line":"|\r"},{"line_number":188,"context_line":"|\r"},{"line_number":189,"context_line":"     5. Images from Glance are pulled down to the local disk of the Ironic conductor servicing the bare metal node.\r"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_ea8c3783","line":186,"updated":"2014-06-23 08:11:32.000000000","message":"As a consequence of this, I think the deployment_steps.png diagram needs a link from Ironic API to the database as well. i.e. https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/node.py#L661","commit_id":"cea73de582784fad8245e64c9cd8f1edbaafac46"},{"author":{"_account_id":8106,"name":"Haomeng,Wang","email":"wanghaomeng@gmail.com","username":"whaom"},"change_message_id":"431155b217c9acc1714915d90769ad613b71763f","unresolved":false,"context_lines":[{"line_number":190,"context_line":"|\r"},{"line_number":191,"context_line":"|\r"},{"line_number":192,"context_line":"     6. Virtual interfaces are plugged in.\r"},{"line_number":193,"context_line":"|\r"},{"line_number":194,"context_line":"|\r"},{"line_number":195,"context_line":"     7. Nova\u0027s ironic driver issues a deploy request via the Ironic API to the Ironic conductor servicing the bare metal node.\r"},{"line_number":196,"context_line":"|\r"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_a4698828","line":193,"updated":"2014-06-23 07:22:32.000000000","message":"And call neutron API to update DHCP port to support PXE/TFTP options.","commit_id":"cea73de582784fad8245e64c9cd8f1edbaafac46"},{"author":{"_account_id":8106,"name":"Haomeng,Wang","email":"wanghaomeng@gmail.com","username":"whaom"},"change_message_id":"431155b217c9acc1714915d90769ad613b71763f","unresolved":false,"context_lines":[{"line_number":201,"context_line":"     9. The IPMI driver issues command to enable network boot of a node and power it on.\r"},{"line_number":202,"context_line":"|\r"},{"line_number":203,"context_line":"|\r"},{"line_number":204,"context_line":"     10. The DHCP boots the deploy ramdisk. The PXE driver actually copies the image over iSCSI to the physical node. It connects to the iSCSI end point, partitions volume, \"dd\" the image and closes the iSCSI connection. The deployment is done.\r"},{"line_number":205,"context_line":"|\r"},{"line_number":206,"context_line":"|\r"},{"line_number":207,"context_line":"     11. The IPMI driver reboots the bare metal node. Note that there are 2 power cycles during bare metal deployment; the first time when powered-on, the images get deployed as mentioned in step 9. The second time as in this case, after the images are deployed, the node is powered up.\r"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_a4d7c8bf","line":204,"updated":"2014-06-23 07:22:32.000000000","message":"And Ironic conductor will switch pxe config to service mode, and notify ramdisk agent the deployment is done.","commit_id":"cea73de582784fad8245e64c9cd8f1edbaafac46"}]}
