)]}'
{"specs/approved/boot-from-ramdisk.rst":[{"author":{"_account_id":14826,"name":"Mark Goddard","email":"markgoddard86@gmail.com","username":"mgoddard"},"change_message_id":"51776035cdfcea7887093a4191909aafca4575fc","unresolved":false,"context_lines":[{"line_number":20,"context_line":""},{"line_number":21,"context_line":"Most of these cases are where an operator is sufficiently knowledgable to"},{"line_number":22,"context_line":"construct a ramdisk and kernel image that can be explicitly booted from"},{"line_number":23,"context_line":"RAM."},{"line_number":24,"context_line":""},{"line_number":25,"context_line":"Problem description"},{"line_number":26,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_d243806b","line":23,"updated":"2018-06-20 10:39:01.000000000","message":"I\u0027m not sure where these things would be added, but some of the reasons for using boot from ramdisk in large HPC environments are:\n\n* fast, scalable booting (skips the deploy image boot and image write phases)\n* ephemeral instances make it easy to return an entire system to a known state\n* diskless compute nodes remove one common failure mode - failures in HPC are expensive when disrupting or even killing a parallel job midway","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"f63a314e9b1b9012c7e563e90a579d241dfa3f16","unresolved":false,"context_lines":[{"line_number":20,"context_line":""},{"line_number":21,"context_line":"Most of these cases are where an operator is sufficiently knowledgable to"},{"line_number":22,"context_line":"construct a ramdisk and kernel image that can be explicitly booted from"},{"line_number":23,"context_line":"RAM."},{"line_number":24,"context_line":""},{"line_number":25,"context_line":"Problem description"},{"line_number":26,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_0ae2b930","line":23,"in_reply_to":"5f7c97a3_2d0805d7","updated":"2018-06-26 01:33:05.000000000","message":"Oh, that is a good point, ad-hoc security booting. I guess a future consideration could be to enable the driver such that it changed the driver to enable specific workflows, such as booting an attestation server registration ramdisk.","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":6133,"name":"yolanda.robla","email":"yroblamo@redhat.com","username":"yolanda.robla"},"change_message_id":"7427286dfc5770ca8e64e98a756285034dd4e238","unresolved":false,"context_lines":[{"line_number":20,"context_line":""},{"line_number":21,"context_line":"Most of these cases are where an operator is sufficiently knowledgable to"},{"line_number":22,"context_line":"construct a ramdisk and kernel image that can be explicitly booted from"},{"line_number":23,"context_line":"RAM."},{"line_number":24,"context_line":""},{"line_number":25,"context_line":"Problem description"},{"line_number":26,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_2d0805d7","line":23,"in_reply_to":"5f7c97a3_d243806b","updated":"2018-06-25 09:02:21.000000000","message":"The secure boot can also get benefit of that right? Booting from ramdisk, to perform the security checks.","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":14826,"name":"Mark Goddard","email":"markgoddard86@gmail.com","username":"mgoddard"},"change_message_id":"51776035cdfcea7887093a4191909aafca4575fc","unresolved":false,"context_lines":[{"line_number":60,"context_line":"  #: The instance can boot based upon any provided"},{"line_number":61,"context_line":"     instance ``kernel`` and ``ramdisk`` as part of"},{"line_number":62,"context_line":"     the requested \"glance\" image to be deployed."},{"line_number":63,"context_line":"     This may also be able to be explicitly defined for direct users."},{"line_number":64,"context_line":"  #: The instance could be booted to a default ``kernel`` and ``ramdisk``"},{"line_number":65,"context_line":"     value that could potentially be stored in ``driver_info``"},{"line_number":66,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_3215149c","line":63,"updated":"2018-06-20 10:39:01.000000000","message":"Just to clarify, this would be a dummy glance image used solely for linking to the ramdisk and kernel?","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":6133,"name":"yolanda.robla","email":"yroblamo@redhat.com","username":"yolanda.robla"},"change_message_id":"7427286dfc5770ca8e64e98a756285034dd4e238","unresolved":false,"context_lines":[{"line_number":60,"context_line":"  #: The instance can boot based upon any provided"},{"line_number":61,"context_line":"     instance ``kernel`` and ``ramdisk`` as part of"},{"line_number":62,"context_line":"     the requested \"glance\" image to be deployed."},{"line_number":63,"context_line":"     This may also be able to be explicitly defined for direct users."},{"line_number":64,"context_line":"  #: The instance could be booted to a default ``kernel`` and ``ramdisk``"},{"line_number":65,"context_line":"     value that could potentially be stored in ``driver_info``"},{"line_number":66,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_8d211950","line":63,"range":{"start_line":63,"start_character":32,"end_line":63,"end_character":33},"updated":"2018-06-25 09:02:21.000000000","message":"We will always use glance to store the images? Or can we think of another options, such to store the images on a remote filesystem and download from there?\nSeems that you are explaining that later, such as boot from NFS, maybe you can add some pointer here.","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"d9b0beb4fce1494aa444b54277fdb3b6755cfbe5","unresolved":false,"context_lines":[{"line_number":60,"context_line":"  #: The instance can boot based upon any provided"},{"line_number":61,"context_line":"     instance ``kernel`` and ``ramdisk`` as part of"},{"line_number":62,"context_line":"     the requested \"glance\" image to be deployed."},{"line_number":63,"context_line":"     This may also be able to be explicitly defined for direct users."},{"line_number":64,"context_line":"  #: The instance could be booted to a default ``kernel`` and ``ramdisk``"},{"line_number":65,"context_line":"     value that could potentially be stored in ``driver_info``"},{"line_number":66,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_61c4eb60","line":63,"in_reply_to":"5f7c97a3_3215149c","updated":"2018-06-22 05:19:49.000000000","message":"If so desired to use with nova, that was kind of what I was thinking.","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"f63a314e9b1b9012c7e563e90a579d241dfa3f16","unresolved":false,"context_lines":[{"line_number":60,"context_line":"  #: The instance can boot based upon any provided"},{"line_number":61,"context_line":"     instance ``kernel`` and ``ramdisk`` as part of"},{"line_number":62,"context_line":"     the requested \"glance\" image to be deployed."},{"line_number":63,"context_line":"     This may also be able to be explicitly defined for direct users."},{"line_number":64,"context_line":"  #: The instance could be booted to a default ``kernel`` and ``ramdisk``"},{"line_number":65,"context_line":"     value that could potentially be stored in ``driver_info``"},{"line_number":66,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_4ac5118d","line":63,"range":{"start_line":63,"start_character":32,"end_line":63,"end_character":33},"in_reply_to":"5f7c97a3_8d211950","updated":"2018-06-26 01:33:05.000000000","message":"I think we would be agnostic in the driver, ultimately supporting whatever the user supplies in the same option as we do presently. URL or glance id.\n\nNFS would make sense for other cases.... I think we would still need to chain load...","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":6133,"name":"yolanda.robla","email":"yroblamo@redhat.com","username":"yolanda.robla"},"change_message_id":"7427286dfc5770ca8e64e98a756285034dd4e238","unresolved":false,"context_lines":[{"line_number":61,"context_line":"     instance ``kernel`` and ``ramdisk`` as part of"},{"line_number":62,"context_line":"     the requested \"glance\" image to be deployed."},{"line_number":63,"context_line":"     This may also be able to be explicitly defined for direct users."},{"line_number":64,"context_line":"  #: The instance could be booted to a default ``kernel`` and ``ramdisk``"},{"line_number":65,"context_line":"     value that could potentially be stored in ``driver_info``"},{"line_number":66,"context_line":""},{"line_number":67,"context_line":"Building this functionality would allow us to also quickly adapt the"}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_4d1781ef","line":64,"range":{"start_line":64,"start_character":49,"end_line":64,"end_character":55},"updated":"2018-06-25 09:02:21.000000000","message":"Always kernel and ramdisk? If we boot with whole disk images, we just need a single image","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"f63a314e9b1b9012c7e563e90a579d241dfa3f16","unresolved":false,"context_lines":[{"line_number":61,"context_line":"     instance ``kernel`` and ``ramdisk`` as part of"},{"line_number":62,"context_line":"     the requested \"glance\" image to be deployed."},{"line_number":63,"context_line":"     This may also be able to be explicitly defined for direct users."},{"line_number":64,"context_line":"  #: The instance could be booted to a default ``kernel`` and ``ramdisk``"},{"line_number":65,"context_line":"     value that could potentially be stored in ``driver_info``"},{"line_number":66,"context_line":""},{"line_number":67,"context_line":"Building this functionality would allow us to also quickly adapt the"}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_aacaa59c","line":64,"range":{"start_line":64,"start_character":49,"end_line":64,"end_character":55},"in_reply_to":"5f7c97a3_4d1781ef","updated":"2018-06-26 01:33:05.000000000","message":"I think we\u0027re confusing the use there. Although if something is bios style booting like the kernel would from a floppy, then it would just be one file. We could iterate and support that with ipxe I guess","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":14826,"name":"Mark Goddard","email":"markgoddard86@gmail.com","username":"mgoddard"},"change_message_id":"51776035cdfcea7887093a4191909aafca4575fc","unresolved":false,"context_lines":[{"line_number":62,"context_line":"     the requested \"glance\" image to be deployed."},{"line_number":63,"context_line":"     This may also be able to be explicitly defined for direct users."},{"line_number":64,"context_line":"  #: The instance could be booted to a default ``kernel`` and ``ramdisk``"},{"line_number":65,"context_line":"     value that could potentially be stored in ``driver_info``"},{"line_number":66,"context_line":""},{"line_number":67,"context_line":"Building this functionality would allow us to also quickly adapt the"},{"line_number":68,"context_line":"support to enable Ceph RBD based booting as well as ramdisk booting"}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_f52226a3","line":65,"updated":"2018-06-20 10:39:01.000000000","message":"This seems a little unusual. Is there a use case behind this?","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"d9b0beb4fce1494aa444b54277fdb3b6755cfbe5","unresolved":false,"context_lines":[{"line_number":62,"context_line":"     the requested \"glance\" image to be deployed."},{"line_number":63,"context_line":"     This may also be able to be explicitly defined for direct users."},{"line_number":64,"context_line":"  #: The instance could be booted to a default ``kernel`` and ``ramdisk``"},{"line_number":65,"context_line":"     value that could potentially be stored in ``driver_info``"},{"line_number":66,"context_line":""},{"line_number":67,"context_line":"Building this functionality would allow us to also quickly adapt the"},{"line_number":68,"context_line":"support to enable Ceph RBD based booting as well as ramdisk booting"}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_21a033a5","line":65,"in_reply_to":"5f7c97a3_f52226a3","updated":"2018-06-22 05:19:49.000000000","message":"I guess it is kind of unusual, it was a logical progression in my brain.","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":14826,"name":"Mark Goddard","email":"markgoddard86@gmail.com","username":"mgoddard"},"change_message_id":"51776035cdfcea7887093a4191909aafca4575fc","unresolved":false,"context_lines":[{"line_number":65,"context_line":"     value that could potentially be stored in ``driver_info``"},{"line_number":66,"context_line":""},{"line_number":67,"context_line":"Building this functionality would allow us to also quickly adapt the"},{"line_number":68,"context_line":"support to enable Ceph RBD based booting as well as ramdisk booting"},{"line_number":69,"context_line":"to an NFS share. A feature that some operators have expressed desire for."},{"line_number":70,"context_line":""},{"line_number":71,"context_line":"Alternatives"}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_d554ca00","line":68,"updated":"2018-06-20 10:39:01.000000000","message":"Wouldn\u0027t we want to build RBD support into the existing deploy drivers, so that it works transparently like iSCSI volumes? I don\u0027t know the details on this, although it\u0027s definitely something we\u0027re interested in.","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":14826,"name":"Mark Goddard","email":"markgoddard86@gmail.com","username":"mgoddard"},"change_message_id":"a14badbf50d1073131339456380564bd75bfa16b","unresolved":false,"context_lines":[{"line_number":65,"context_line":"     value that could potentially be stored in ``driver_info``"},{"line_number":66,"context_line":""},{"line_number":67,"context_line":"Building this functionality would allow us to also quickly adapt the"},{"line_number":68,"context_line":"support to enable Ceph RBD based booting as well as ramdisk booting"},{"line_number":69,"context_line":"to an NFS share. A feature that some operators have expressed desire for."},{"line_number":70,"context_line":""},{"line_number":71,"context_line":"Alternatives"}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_bded43b9","line":68,"in_reply_to":"5f7c97a3_a1ce8380","updated":"2018-06-22 11:02:37.000000000","message":"Ah, that\u0027s true - there is presumably no RBD support in iPXE so you\u0027d need to first boot a ramdisk. Hopefully the required kernel parameters are distro-agnostic.","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"f63a314e9b1b9012c7e563e90a579d241dfa3f16","unresolved":false,"context_lines":[{"line_number":65,"context_line":"     value that could potentially be stored in ``driver_info``"},{"line_number":66,"context_line":""},{"line_number":67,"context_line":"Building this functionality would allow us to also quickly adapt the"},{"line_number":68,"context_line":"support to enable Ceph RBD based booting as well as ramdisk booting"},{"line_number":69,"context_line":"to an NFS share. A feature that some operators have expressed desire for."},{"line_number":70,"context_line":""},{"line_number":71,"context_line":"Alternatives"}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_0a5079d4","line":68,"in_reply_to":"5f7c97a3_bded43b9","updated":"2018-06-26 01:33:05.000000000","message":"sadly they are not :\\ One of the reasons we punted on it to begin with. Luckily the dracut based options should be uniform.","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"d9b0beb4fce1494aa444b54277fdb3b6755cfbe5","unresolved":false,"context_lines":[{"line_number":65,"context_line":"     value that could potentially be stored in ``driver_info``"},{"line_number":66,"context_line":""},{"line_number":67,"context_line":"Building this functionality would allow us to also quickly adapt the"},{"line_number":68,"context_line":"support to enable Ceph RBD based booting as well as ramdisk booting"},{"line_number":69,"context_line":"to an NFS share. A feature that some operators have expressed desire for."},{"line_number":70,"context_line":""},{"line_number":71,"context_line":"Alternatives"}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_a1ce8380","line":68,"in_reply_to":"5f7c97a3_d554ca00","updated":"2018-06-22 05:19:49.000000000","message":"So we can do that, but the RDB volume support will require a kernel/ramdisk to be set.","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":14826,"name":"Mark Goddard","email":"markgoddard86@gmail.com","username":"mgoddard"},"change_message_id":"51776035cdfcea7887093a4191909aafca4575fc","unresolved":false,"context_lines":[{"line_number":74,"context_line":"We could consider building logic to handle and support this into the standard"},{"line_number":75,"context_line":"deployment workflow, however given this is a sufficient and specialized use"},{"line_number":76,"context_line":"case, that it might be completely unnecessary as it would not be used in most"},{"line_number":77,"context_line":"cases."},{"line_number":78,"context_line":""},{"line_number":79,"context_line":"Data model impact"},{"line_number":80,"context_line":"-----------------"}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_359c1ebe","line":77,"updated":"2018-06-20 10:39:01.000000000","message":"From a naming perspective, neither iscsi nor agent fits the bill, so a new deploy interface makes sense.","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":14826,"name":"Mark Goddard","email":"markgoddard86@gmail.com","username":"mgoddard"},"change_message_id":"51776035cdfcea7887093a4191909aafca4575fc","unresolved":false,"context_lines":[{"line_number":132,"context_line":"being able to be setup by an operator where it may be necessary to keep the"},{"line_number":133,"context_line":"deployed kernel/ramdisk potentially for an elongated period of time."},{"line_number":134,"context_line":""},{"line_number":135,"context_line":"Ironic already does this with instances that are netboot."},{"line_number":136,"context_line":""},{"line_number":137,"context_line":"Other end user impact"},{"line_number":138,"context_line":"---------------------"}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_72414ca6","line":135,"updated":"2018-06-20 10:39:01.000000000","message":"It also requires access to the same TFTP/HTTP server used for provisioning, which is normally only accessible from secure networks. This is also true of netboot, but AFAIK netboot is not recommended.","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"d9b0beb4fce1494aa444b54277fdb3b6755cfbe5","unresolved":false,"context_lines":[{"line_number":132,"context_line":"being able to be setup by an operator where it may be necessary to keep the"},{"line_number":133,"context_line":"deployed kernel/ramdisk potentially for an elongated period of time."},{"line_number":134,"context_line":""},{"line_number":135,"context_line":"Ironic already does this with instances that are netboot."},{"line_number":136,"context_line":""},{"line_number":137,"context_line":"Other end user impact"},{"line_number":138,"context_line":"---------------------"}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_a193e35b","line":135,"in_reply_to":"5f7c97a3_72414ca6","updated":"2018-06-22 05:19:49.000000000","message":"Great point, will add.","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":14826,"name":"Mark Goddard","email":"markgoddard86@gmail.com","username":"mgoddard"},"change_message_id":"51776035cdfcea7887093a4191909aafca4575fc","unresolved":false,"context_lines":[{"line_number":138,"context_line":"---------------------"},{"line_number":139,"context_line":""},{"line_number":140,"context_line":"Operators which choose to use this feature may wish to put in place"},{"line_number":141,"context_line":"specialized network controls to facilitate the machines network booting."},{"line_number":142,"context_line":""},{"line_number":143,"context_line":"Each deployment and case will be different, and without post-implementation"},{"line_number":144,"context_line":"information, we will be unable to determine if there is a standard that can"}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_b5f10ee1","line":141,"updated":"2018-06-20 10:39:01.000000000","message":"We\u0027ve been struggling to work out how boot from volume would work in a tenant networking environment, and this falls into some of the same issues. At least we don\u0027t need to reach the storage network in this case.\n\nIdeally we\u0027d do any required networking in hardware rather than software (particularly true for iSCSI traffic).\n\nSome parallel file systems dedicate some nodes to shuffling traffic between networks (sometimes with different technologies, e.g. InfiniBand \u003c-\u003e Ethernet, see LNET routers [1] for an example). It could be possible to do something similar here with nodes straddling tenant networks and provisioning and storage networks.\n\nPerhaps a reference design could include a neutron router routing between these networks.\n\n[1] http://wiki.lustre.org/LNET_Router","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"d9b0beb4fce1494aa444b54277fdb3b6755cfbe5","unresolved":false,"context_lines":[{"line_number":138,"context_line":"---------------------"},{"line_number":139,"context_line":""},{"line_number":140,"context_line":"Operators which choose to use this feature may wish to put in place"},{"line_number":141,"context_line":"specialized network controls to facilitate the machines network booting."},{"line_number":142,"context_line":""},{"line_number":143,"context_line":"Each deployment and case will be different, and without post-implementation"},{"line_number":144,"context_line":"information, we will be unable to determine if there is a standard that can"}],"source_content_type":"text/x-rst","patch_set":3,"id":"5f7c97a3_61994b78","line":141,"in_reply_to":"5f7c97a3_b5f10ee1","updated":"2018-06-22 05:19:49.000000000","message":"Added this thought to the processing queue in my brain. Will try to respond in the next day or two.","commit_id":"a0941c5e1715ea24edb3ea9769de09e99f5ebd55"},{"author":{"_account_id":6133,"name":"yolanda.robla","email":"yroblamo@redhat.com","username":"yolanda.robla"},"change_message_id":"5da4e36342e9035063f87f417f1be30831be16be","unresolved":false,"context_lines":[{"line_number":79,"context_line":""},{"line_number":80,"context_line":"Building this functionality would allow us to also quickly adapt the"},{"line_number":81,"context_line":"support to enable Ceph RBD based booting as well as ramdisk booting"},{"line_number":82,"context_line":"to an NFS share serving as the root device. Both feature that some"},{"line_number":83,"context_line":"operators have expressed desire for. Ceph RBD support would need to"},{"line_number":84,"context_line":"be in the underlying ``pxe`` boot interface, as the ramdisk interface"},{"line_number":85,"context_line":"overlays the ``pxe`` interface."}],"source_content_type":"text/x-rst","patch_set":8,"id":"5f7c97a3_3857a504","line":82,"range":{"start_line":82,"start_character":49,"end_line":82,"end_character":56},"updated":"2018-06-27 14:49:39.000000000","message":"features","commit_id":"a54b9d1d1f437f87b1327f3e9e75693b935c2db6"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"6001faf47631e25269501c876a1440031a296970","unresolved":false,"context_lines":[{"line_number":58,"context_line":"classes, in order to minimize complexity and allow for cleaning workflow"},{"line_number":59,"context_line":"and related code to remain in one location."},{"line_number":60,"context_line":""},{"line_number":61,"context_line":"Mechanism wise, if the ``ramdisk`` deployment interface is set:"},{"line_number":62,"context_line":" * The instance can boot based upon any provided"},{"line_number":63,"context_line":"   instance ``kernel`` and ``ramdisk`` as part of"},{"line_number":64,"context_line":"   the requested \"glance\" image to be deployed."}],"source_content_type":"text/x-rst","patch_set":9,"id":"5f7c97a3_56b18444","line":61,"updated":"2018-07-17 12:45:59.000000000","message":"nit: this list has formatting issues, see the rendered version","commit_id":"39f258671066c0968b57538bf19d27fff59f2113"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"6001faf47631e25269501c876a1440031a296970","unresolved":false,"context_lines":[{"line_number":62,"context_line":" * The instance can boot based upon any provided"},{"line_number":63,"context_line":"   instance ``kernel`` and ``ramdisk`` as part of"},{"line_number":64,"context_line":"   the requested \"glance\" image to be deployed."},{"line_number":65,"context_line":"   Effectively any partition image or disk image would be"},{"line_number":66,"context_line":"   ignored."},{"line_number":67,"context_line":""},{"line_number":68,"context_line":" * The same ``kernel`` and ``ramdisk`` parameters that are used"}],"source_content_type":"text/x-rst","patch_set":9,"id":"5f7c97a3_96d95c85","line":65,"range":{"start_line":65,"start_character":38,"end_line":65,"end_character":48},"updated":"2018-07-17 12:45:59.000000000","message":"nit: whole-disk images do not have kernel/ramdisk specified","commit_id":"39f258671066c0968b57538bf19d27fff59f2113"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"cca24e5ea8b02b095ce71482d5e06ada5fd0b21c","unresolved":false,"context_lines":[{"line_number":62,"context_line":" * The instance can boot based upon any provided"},{"line_number":63,"context_line":"   instance ``kernel`` and ``ramdisk`` as part of"},{"line_number":64,"context_line":"   the requested \"glance\" image to be deployed."},{"line_number":65,"context_line":"   Effectively any partition image or disk image would be"},{"line_number":66,"context_line":"   ignored."},{"line_number":67,"context_line":""},{"line_number":68,"context_line":" * The same ``kernel`` and ``ramdisk`` parameters that are used"}],"source_content_type":"text/x-rst","patch_set":9,"id":"5f7c97a3_3d3845d2","line":65,"range":{"start_line":65,"start_character":38,"end_line":65,"end_character":48},"in_reply_to":"5f7c97a3_4d524b38","updated":"2018-07-18 14:08:18.000000000","message":"Well, we distinguish partition and whole disk images by presence of kernel/ramdisk. Or am I confusing something?","commit_id":"39f258671066c0968b57538bf19d27fff59f2113"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"15ba6bc88fc32dc53dbc68ae5f0377d6699a374b","unresolved":false,"context_lines":[{"line_number":62,"context_line":" * The instance can boot based upon any provided"},{"line_number":63,"context_line":"   instance ``kernel`` and ``ramdisk`` as part of"},{"line_number":64,"context_line":"   the requested \"glance\" image to be deployed."},{"line_number":65,"context_line":"   Effectively any partition image or disk image would be"},{"line_number":66,"context_line":"   ignored."},{"line_number":67,"context_line":""},{"line_number":68,"context_line":" * The same ``kernel`` and ``ramdisk`` parameters that are used"}],"source_content_type":"text/x-rst","patch_set":9,"id":"5f7c97a3_4d524b38","line":65,"range":{"start_line":65,"start_character":38,"end_line":65,"end_character":48},"in_reply_to":"5f7c97a3_96d95c85","updated":"2018-07-18 14:06:19.000000000","message":"They do not need to I guess is the way to put it. I guess it would be better to express as that any image contents that consist of a filesystem would be ignored.","commit_id":"39f258671066c0968b57538bf19d27fff59f2113"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"6001faf47631e25269501c876a1440031a296970","unresolved":false,"context_lines":[{"line_number":69,"context_line":"    for a \"glance\" image based deployment could be re-used to contain"},{"line_number":70,"context_line":"    a URL."},{"line_number":71,"context_line":""},{"line_number":72,"context_line":"    * We `should` consider extending the accepted URL format"},{"line_number":73,"context_line":"      to include \"nfs://${server_ip}{$path}{$file}\", which"},{"line_number":74,"context_line":"      would enable direct boot to NFS hosted kernel and ramdisk."},{"line_number":75,"context_line":"      This would not consist of ``nfsroot`` scenarios."}],"source_content_type":"text/x-rst","patch_set":9,"id":"5f7c97a3_d6c4b496","line":72,"range":{"start_line":72,"start_character":9,"end_line":72,"end_character":26},"updated":"2018-07-17 12:45:59.000000000","message":"does it mean in the future or as part of this spec? more details are needed if the latter","commit_id":"39f258671066c0968b57538bf19d27fff59f2113"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"cca24e5ea8b02b095ce71482d5e06ada5fd0b21c","unresolved":false,"context_lines":[{"line_number":69,"context_line":"    for a \"glance\" image based deployment could be re-used to contain"},{"line_number":70,"context_line":"    a URL."},{"line_number":71,"context_line":""},{"line_number":72,"context_line":"    * We `should` consider extending the accepted URL format"},{"line_number":73,"context_line":"      to include \"nfs://${server_ip}{$path}{$file}\", which"},{"line_number":74,"context_line":"      would enable direct boot to NFS hosted kernel and ramdisk."},{"line_number":75,"context_line":"      This would not consist of ``nfsroot`` scenarios."}],"source_content_type":"text/x-rst","patch_set":9,"id":"5f7c97a3_3d3da5bb","line":72,"range":{"start_line":72,"start_character":9,"end_line":72,"end_character":26},"in_reply_to":"5f7c97a3_0d60d30c","updated":"2018-07-18 14:08:18.000000000","message":"Then let\u0027s make it clear that this spec does not propose it. Maybe a separate section \"Potential future extension\"?","commit_id":"39f258671066c0968b57538bf19d27fff59f2113"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"15ba6bc88fc32dc53dbc68ae5f0377d6699a374b","unresolved":false,"context_lines":[{"line_number":69,"context_line":"    for a \"glance\" image based deployment could be re-used to contain"},{"line_number":70,"context_line":"    a URL."},{"line_number":71,"context_line":""},{"line_number":72,"context_line":"    * We `should` consider extending the accepted URL format"},{"line_number":73,"context_line":"      to include \"nfs://${server_ip}{$path}{$file}\", which"},{"line_number":74,"context_line":"      would enable direct boot to NFS hosted kernel and ramdisk."},{"line_number":75,"context_line":"      This would not consist of ``nfsroot`` scenarios."}],"source_content_type":"text/x-rst","patch_set":9,"id":"5f7c97a3_0d60d30c","line":72,"range":{"start_line":72,"start_character":9,"end_line":72,"end_character":26},"in_reply_to":"5f7c97a3_d6c4b496","updated":"2018-07-18 14:06:19.000000000","message":"Future","commit_id":"39f258671066c0968b57538bf19d27fff59f2113"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"6001faf47631e25269501c876a1440031a296970","unresolved":false,"context_lines":[{"line_number":70,"context_line":"    a URL."},{"line_number":71,"context_line":""},{"line_number":72,"context_line":"    * We `should` consider extending the accepted URL format"},{"line_number":73,"context_line":"      to include \"nfs://${server_ip}{$path}{$file}\", which"},{"line_number":74,"context_line":"      would enable direct boot to NFS hosted kernel and ramdisk."},{"line_number":75,"context_line":"      This would not consist of ``nfsroot`` scenarios."},{"line_number":76,"context_line":""}],"source_content_type":"text/x-rst","patch_set":9,"id":"5f7c97a3_568ae464","line":73,"updated":"2018-07-17 12:45:59.000000000","message":"nit: `` around URL","commit_id":"39f258671066c0968b57538bf19d27fff59f2113"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"90d40310c8db2b8c3e6aa84b34bf90a566165ff5","unresolved":false,"context_lines":[{"line_number":83,"context_line":"operators have expressed desire for. Ceph RBD support would need to"},{"line_number":84,"context_line":"be in the underlying ``pxe`` boot interface, as the ramdisk interface"},{"line_number":85,"context_line":"overlays the ``pxe`` interface."},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"Alternatives"},{"line_number":88,"context_line":"------------"},{"line_number":89,"context_line":""}],"source_content_type":"text/x-rst","patch_set":9,"id":"5f7c97a3_16e6ac4b","line":86,"updated":"2018-07-17 12:44:56.000000000","message":"This is great for ironic users, how do we solve for nova users with config drives?","commit_id":"39f258671066c0968b57538bf19d27fff59f2113"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"6001faf47631e25269501c876a1440031a296970","unresolved":false,"context_lines":[{"line_number":83,"context_line":"operators have expressed desire for. Ceph RBD support would need to"},{"line_number":84,"context_line":"be in the underlying ``pxe`` boot interface, as the ramdisk interface"},{"line_number":85,"context_line":"overlays the ``pxe`` interface."},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"Alternatives"},{"line_number":88,"context_line":"------------"},{"line_number":89,"context_line":""}],"source_content_type":"text/x-rst","patch_set":9,"id":"5f7c97a3_b691f89c","line":86,"updated":"2018-07-17 12:45:59.000000000","message":"we need to state: 1. what to do with configdrive, 2. caveats around networking","commit_id":"39f258671066c0968b57538bf19d27fff59f2113"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"15ba6bc88fc32dc53dbc68ae5f0377d6699a374b","unresolved":false,"context_lines":[{"line_number":83,"context_line":"operators have expressed desire for. Ceph RBD support would need to"},{"line_number":84,"context_line":"be in the underlying ``pxe`` boot interface, as the ramdisk interface"},{"line_number":85,"context_line":"overlays the ``pxe`` interface."},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"Alternatives"},{"line_number":88,"context_line":"------------"},{"line_number":89,"context_line":""}],"source_content_type":"text/x-rst","patch_set":9,"id":"5f7c97a3_6d4f078c","line":86,"in_reply_to":"5f7c97a3_b691f89c","updated":"2018-07-18 14:06:19.000000000","message":"++, will do in next revision.","commit_id":"39f258671066c0968b57538bf19d27fff59f2113"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"34485c81fafbb20c11298bd34b5b59fd0790c041","unresolved":false,"context_lines":[{"line_number":46,"context_line":"Proposed change"},{"line_number":47,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":48,"context_line":""},{"line_number":49,"context_line":"We propose to build a ``ramdisk`` deployment interface built upon our"},{"line_number":50,"context_line":"existing ramdisk code and facilities. This would allow us to have a"},{"line_number":51,"context_line":"jump start on working functionality and help us better identify where"},{"line_number":52,"context_line":"refactoring must be required."}],"source_content_type":"text/x-rst","patch_set":10,"id":"5f7c97a3_8480e7f7","line":49,"updated":"2018-07-20 14:34:06.000000000","message":"nit: the canonical name is \"deploy interface\"","commit_id":"7cd6df790e104d1b509deb1df8cbc2d7070006a4"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"dff63cb05d168f70796c3dd93c60e21d74f5ddb2","unresolved":false,"context_lines":[{"line_number":46,"context_line":"Proposed change"},{"line_number":47,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":48,"context_line":""},{"line_number":49,"context_line":"We propose to build a ``ramdisk`` deployment interface built upon our"},{"line_number":50,"context_line":"existing ramdisk code and facilities. This would allow us to have a"},{"line_number":51,"context_line":"jump start on working functionality and help us better identify where"},{"line_number":52,"context_line":"refactoring must be required."}],"source_content_type":"text/x-rst","patch_set":10,"id":"5f7c97a3_89dbfb4c","line":49,"in_reply_to":"5f7c97a3_8480e7f7","updated":"2018-07-23 12:47:53.000000000","message":"Done","commit_id":"7cd6df790e104d1b509deb1df8cbc2d7070006a4"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"34485c81fafbb20c11298bd34b5b59fd0790c041","unresolved":false,"context_lines":[{"line_number":63,"context_line":"* The instance can boot based upon any provided"},{"line_number":64,"context_line":"  instance ``kernel`` and ``ramdisk`` as part of"},{"line_number":65,"context_line":"  the requested \"glance\" image to be deployed."},{"line_number":66,"context_line":"  Effectively any image contents would be ignored."},{"line_number":67,"context_line":""},{"line_number":68,"context_line":"* The same ``kernel`` and ``ramdisk`` parameters that are used"},{"line_number":69,"context_line":"  for a \"glance\" image based deployment could be re-used to contain"}],"source_content_type":"text/x-rst","patch_set":10,"id":"5f7c97a3_247b1329","line":66,"updated":"2018-07-20 14:34:06.000000000","message":"I guess boot_option is ignored as well?","commit_id":"7cd6df790e104d1b509deb1df8cbc2d7070006a4"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"dff63cb05d168f70796c3dd93c60e21d74f5ddb2","unresolved":false,"context_lines":[{"line_number":63,"context_line":"* The instance can boot based upon any provided"},{"line_number":64,"context_line":"  instance ``kernel`` and ``ramdisk`` as part of"},{"line_number":65,"context_line":"  the requested \"glance\" image to be deployed."},{"line_number":66,"context_line":"  Effectively any image contents would be ignored."},{"line_number":67,"context_line":""},{"line_number":68,"context_line":"* The same ``kernel`` and ``ramdisk`` parameters that are used"},{"line_number":69,"context_line":"  for a \"glance\" image based deployment could be re-used to contain"}],"source_content_type":"text/x-rst","patch_set":10,"id":"5f7c97a3_29e0a7f9","line":66,"in_reply_to":"5f7c97a3_247b1329","updated":"2018-07-23 12:47:53.000000000","message":"Done","commit_id":"7cd6df790e104d1b509deb1df8cbc2d7070006a4"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"70f1b068bcd52d2a0cf8bc45b4f8f65a1b1d7511","unresolved":false,"context_lines":[{"line_number":63,"context_line":"* The instance can boot based upon any provided"},{"line_number":64,"context_line":"  instance ``kernel`` and ``ramdisk`` as part of"},{"line_number":65,"context_line":"  the requested \"glance\" image to be deployed."},{"line_number":66,"context_line":"  Effectively any image contents would be ignored."},{"line_number":67,"context_line":""},{"line_number":68,"context_line":"* The same ``kernel`` and ``ramdisk`` parameters that are used"},{"line_number":69,"context_line":"  for a \"glance\" image based deployment could be re-used to contain"}],"source_content_type":"text/x-rst","patch_set":10,"id":"5f7c97a3_c6eec289","line":66,"in_reply_to":"5f7c97a3_247b1329","updated":"2018-07-23 12:29:49.000000000","message":"Overrides if set.","commit_id":"7cd6df790e104d1b509deb1df8cbc2d7070006a4"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"34485c81fafbb20c11298bd34b5b59fd0790c041","unresolved":false,"context_lines":[{"line_number":83,"context_line":"* Network interface plugging would be included and supported, however"},{"line_number":84,"context_line":"  users must explicitly allow access to the TFTP or HTTP endpoint for"},{"line_number":85,"context_line":"  booting the bare metal node. Typically deployments do not allow tenant"},{"line_number":86,"context_line":"  networks are not permitted access to these endpoints."},{"line_number":87,"context_line":""},{"line_number":88,"context_line":"* Configuration drives would be ignored by ironic, and users would be"},{"line_number":89,"context_line":"  advised to make use of the metadata service."}],"source_content_type":"text/x-rst","patch_set":10,"id":"5f7c97a3_44764f41","line":86,"updated":"2018-07-20 14:34:06.000000000","message":"nit: too many verbs in this sentence?","commit_id":"7cd6df790e104d1b509deb1df8cbc2d7070006a4"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"dff63cb05d168f70796c3dd93c60e21d74f5ddb2","unresolved":false,"context_lines":[{"line_number":83,"context_line":"* Network interface plugging would be included and supported, however"},{"line_number":84,"context_line":"  users must explicitly allow access to the TFTP or HTTP endpoint for"},{"line_number":85,"context_line":"  booting the bare metal node. Typically deployments do not allow tenant"},{"line_number":86,"context_line":"  networks are not permitted access to these endpoints."},{"line_number":87,"context_line":""},{"line_number":88,"context_line":"* Configuration drives would be ignored by ironic, and users would be"},{"line_number":89,"context_line":"  advised to make use of the metadata service."}],"source_content_type":"text/x-rst","patch_set":10,"id":"5f7c97a3_69ca1f1a","line":86,"in_reply_to":"5f7c97a3_44764f41","updated":"2018-07-23 12:47:53.000000000","message":"Rewriting the first half since that should convey the message.","commit_id":"7cd6df790e104d1b509deb1df8cbc2d7070006a4"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"34485c81fafbb20c11298bd34b5b59fd0790c041","unresolved":false,"context_lines":[{"line_number":90,"context_line":""},{"line_number":91,"context_line":"  * Operators who directly deploy nodes using ironic may need to pass"},{"line_number":92,"context_line":"    additional kernel command line arguments to the node being \"deployed\""},{"line_number":93,"context_line":"    via a ramdisk. In this case, a ``/instance_info/additional_kernel_arguments``"},{"line_number":94,"context_line":"    field will be accepted to allow those operators to pass information to the"},{"line_number":95,"context_line":"    instance."},{"line_number":96,"context_line":""}],"source_content_type":"text/x-rst","patch_set":10,"id":"5f7c97a3_e470bb43","line":93,"updated":"2018-07-20 14:34:06.000000000","message":"if this is only interpreted by this interface, won\u0027t it make sense to put it int driver_info? what I\u0027m trying to avoid is people thinking that additional_kernel_arguments work for all deploy methods (incl. final instances). Or do we want to allow passing them to IPA for all deploy interfaces? if yes, maybe call it ramdisk_kernel_argument to make it clear that deployed instances are not affected?","commit_id":"7cd6df790e104d1b509deb1df8cbc2d7070006a4"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"70f1b068bcd52d2a0cf8bc45b4f8f65a1b1d7511","unresolved":false,"context_lines":[{"line_number":90,"context_line":""},{"line_number":91,"context_line":"  * Operators who directly deploy nodes using ironic may need to pass"},{"line_number":92,"context_line":"    additional kernel command line arguments to the node being \"deployed\""},{"line_number":93,"context_line":"    via a ramdisk. In this case, a ``/instance_info/additional_kernel_arguments``"},{"line_number":94,"context_line":"    field will be accepted to allow those operators to pass information to the"},{"line_number":95,"context_line":"    instance."},{"line_number":96,"context_line":""}],"source_content_type":"text/x-rst","patch_set":10,"id":"5f7c97a3_e6a41e4f","line":93,"in_reply_to":"5f7c97a3_e470bb43","updated":"2018-07-23 12:29:49.000000000","message":"I think it would make sense to put it in instance info for if we split out baremetal admin and had a baremetal user... the user would not be able to access the driver_info field or make changes to it.  +1 to ramdisk_kernel_argument.","commit_id":"7cd6df790e104d1b509deb1df8cbc2d7070006a4"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"dff63cb05d168f70796c3dd93c60e21d74f5ddb2","unresolved":false,"context_lines":[{"line_number":90,"context_line":""},{"line_number":91,"context_line":"  * Operators who directly deploy nodes using ironic may need to pass"},{"line_number":92,"context_line":"    additional kernel command line arguments to the node being \"deployed\""},{"line_number":93,"context_line":"    via a ramdisk. In this case, a ``/instance_info/additional_kernel_arguments``"},{"line_number":94,"context_line":"    field will be accepted to allow those operators to pass information to the"},{"line_number":95,"context_line":"    instance."},{"line_number":96,"context_line":""}],"source_content_type":"text/x-rst","patch_set":10,"id":"5f7c97a3_494c03c9","line":93,"in_reply_to":"5f7c97a3_e6a41e4f","updated":"2018-07-23 12:47:53.000000000","message":"FWIW, technically a crafty operator could use this to pass NFS or Ceph RBD parameters along.... But volume target bsaed would be ideal given that Ceph RBD data gets stored from nova for users that try to BFV.","commit_id":"7cd6df790e104d1b509deb1df8cbc2d7070006a4"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"34485c81fafbb20c11298bd34b5b59fd0790c041","unresolved":false,"context_lines":[{"line_number":99,"context_line":"to an NFS share serving as the root device. Both are features some"},{"line_number":100,"context_line":"operators have expressed desire for. Ceph RBD support would need to"},{"line_number":101,"context_line":"be in the underlying ``pxe`` boot interface, as the ramdisk interface"},{"line_number":102,"context_line":"overlays the ``pxe`` interface."},{"line_number":103,"context_line":""},{"line_number":104,"context_line":"Alternatives"},{"line_number":105,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":10,"id":"5f7c97a3_046cd768","line":102,"updated":"2018-07-20 14:34:06.000000000","message":"/me is curious how it would look","commit_id":"7cd6df790e104d1b509deb1df8cbc2d7070006a4"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"70f1b068bcd52d2a0cf8bc45b4f8f65a1b1d7511","unresolved":false,"context_lines":[{"line_number":99,"context_line":"to an NFS share serving as the root device. Both are features some"},{"line_number":100,"context_line":"operators have expressed desire for. Ceph RBD support would need to"},{"line_number":101,"context_line":"be in the underlying ``pxe`` boot interface, as the ramdisk interface"},{"line_number":102,"context_line":"overlays the ``pxe`` interface."},{"line_number":103,"context_line":""},{"line_number":104,"context_line":"Alternatives"},{"line_number":105,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":10,"id":"5f7c97a3_a68e26d5","line":102,"in_reply_to":"5f7c97a3_046cd768","updated":"2018-07-23 12:29:49.000000000","message":"Additional templates or template complexity. Most of these are just additional command line arguments and the BFV spec calls for information storage, so we can just write out if present and the conditions are set.","commit_id":"7cd6df790e104d1b509deb1df8cbc2d7070006a4"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"34485c81fafbb20c11298bd34b5b59fd0790c041","unresolved":false,"context_lines":[{"line_number":213,"context_line":"Work Items"},{"line_number":214,"context_line":"----------"},{"line_number":215,"context_line":""},{"line_number":216,"context_line":"* Create basic driver"},{"line_number":217,"context_line":"* Create tempest test for said ramdisk booting driver."},{"line_number":218,"context_line":"* Create use documentation."},{"line_number":219,"context_line":""}],"source_content_type":"text/x-rst","patch_set":10,"id":"5f7c97a3_a48c032e","line":216,"updated":"2018-07-20 14:34:06.000000000","message":"nit: s/driver/deploy interface/","commit_id":"7cd6df790e104d1b509deb1df8cbc2d7070006a4"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"34485c81fafbb20c11298bd34b5b59fd0790c041","unresolved":false,"context_lines":[{"line_number":214,"context_line":"----------"},{"line_number":215,"context_line":""},{"line_number":216,"context_line":"* Create basic driver"},{"line_number":217,"context_line":"* Create tempest test for said ramdisk booting driver."},{"line_number":218,"context_line":"* Create use documentation."},{"line_number":219,"context_line":""},{"line_number":220,"context_line":"Dependencies"}],"source_content_type":"text/x-rst","patch_set":10,"id":"5f7c97a3_c48fbf1f","line":217,"updated":"2018-07-20 14:34:06.000000000","message":"ditto","commit_id":"7cd6df790e104d1b509deb1df8cbc2d7070006a4"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"34485c81fafbb20c11298bd34b5b59fd0790c041","unresolved":false,"context_lines":[{"line_number":224,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":225,"context_line":""},{"line_number":226,"context_line":"This seems like a feature that could be easily tested via a tempest scenario"},{"line_number":227,"context_line":"if the driver is available. No additional testing should be required."},{"line_number":228,"context_line":""},{"line_number":229,"context_line":"Upgrades and Backwards Compatibility"},{"line_number":230,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":10,"id":"5f7c97a3_64926b4c","line":227,"updated":"2018-07-20 14:34:06.000000000","message":"via the standalone job?","commit_id":"7cd6df790e104d1b509deb1df8cbc2d7070006a4"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"dff63cb05d168f70796c3dd93c60e21d74f5ddb2","unresolved":false,"context_lines":[{"line_number":224,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":225,"context_line":""},{"line_number":226,"context_line":"This seems like a feature that could be easily tested via a tempest scenario"},{"line_number":227,"context_line":"if the driver is available. No additional testing should be required."},{"line_number":228,"context_line":""},{"line_number":229,"context_line":"Upgrades and Backwards Compatibility"},{"line_number":230,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":10,"id":"5f7c97a3_69077f72","line":227,"in_reply_to":"5f7c97a3_64926b4c","updated":"2018-07-23 12:47:53.000000000","message":"That is kind of what I was thinking. It is definitely a use case more geared to someone directly interacting with ironic, so I think if we test it as such, then we wouldn\u0027t really need to worry too much.","commit_id":"7cd6df790e104d1b509deb1df8cbc2d7070006a4"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"f413f80f4bb081ac95a684b667d7e0a6cf2904fc","unresolved":false,"context_lines":[{"line_number":10,"context_line":""},{"line_number":11,"context_line":"https://storyboard.openstack.org/#!/story/1753842"},{"line_number":12,"context_line":""},{"line_number":13,"context_line":"While Ironic, by its very operational nature supports booting from a ramdisk,"},{"line_number":14,"context_line":"we do not empower or enable operators who wish to utilize instances that boot"},{"line_number":15,"context_line":"from ramdisks."},{"line_number":16,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_bbe5c6dd","line":13,"updated":"2018-07-24 08:43:26.000000000","message":"nit: /me wonders if a comma is needed after \"nature\" or not needed befure \"by\"","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":6618,"name":"Ruby Loo","email":"opensrloo@gmail.com","username":"rloo"},"change_message_id":"d3667be04c0c90fb39e01bbe1f9cea67f5adcdcc","unresolved":false,"context_lines":[{"line_number":10,"context_line":""},{"line_number":11,"context_line":"https://storyboard.openstack.org/#!/story/1753842"},{"line_number":12,"context_line":""},{"line_number":13,"context_line":"While Ironic, by its very operational nature supports booting from a ramdisk,"},{"line_number":14,"context_line":"we do not empower or enable operators who wish to utilize instances that boot"},{"line_number":15,"context_line":"from ramdisks."},{"line_number":16,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_3bcc4dbd","line":13,"in_reply_to":"5f7c97a3_bbe5c6dd","updated":"2018-07-24 14:17:00.000000000","message":"I don\u0027t know if it is \u0027needed\u0027 (I was brought up that it is though), but it would be better to have a comma after \u0027nature\u0027.\n\nI think removing the comma would make it that much harder to parse the sentence.","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":6618,"name":"Ruby Loo","email":"opensrloo@gmail.com","username":"rloo"},"change_message_id":"a173a646bc6803bd0f96af27053501923af9ecd0","unresolved":false,"context_lines":[{"line_number":39,"context_line":"  similar with boot from volume, and if we ever support RBD booting,"},{"line_number":40,"context_line":"  we would functionally require this exact feature. What delineates"},{"line_number":41,"context_line":"  this from the Boot from Volume iPXE chain loading scenarios and"},{"line_number":42,"context_line":"  virtual media based booting from a remote volume, is that these"},{"line_number":43,"context_line":"  are ultimately reliant upon back-end block storage, which is"},{"line_number":44,"context_line":"  not always needed or desired for all cases."},{"line_number":45,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_32a57002","line":42,"updated":"2018-07-23 14:38:06.000000000","message":"nit, by \u0027these\u0027 you mean the latter, right?","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"f413f80f4bb081ac95a684b667d7e0a6cf2904fc","unresolved":false,"context_lines":[{"line_number":52,"context_line":"refactoring must be required."},{"line_number":53,"context_line":""},{"line_number":54,"context_line":"The ``ramdisk`` deployment driver interface would be an opt-in interface"},{"line_number":55,"context_line":"which focuses on supporting ramdisk booting cases."},{"line_number":56,"context_line":""},{"line_number":57,"context_line":"The ``ramdisk`` deployment driver would also loop in the agent interface"},{"line_number":58,"context_line":"classes, in order to minimize complexity and allow for cleaning workflow"}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_5bca7247","line":55,"updated":"2018-07-24 08:43:26.000000000","message":"it\u0027s not clear from the spec whether this deploy interface will be enabled by default in enabled_deploy_interfaces and GenericHardware.supported_deploy_interfaces. Since it does not have external dependencies I assume yes to both?","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":6618,"name":"Ruby Loo","email":"opensrloo@gmail.com","username":"rloo"},"change_message_id":"d3667be04c0c90fb39e01bbe1f9cea67f5adcdcc","unresolved":false,"context_lines":[{"line_number":52,"context_line":"refactoring must be required."},{"line_number":53,"context_line":""},{"line_number":54,"context_line":"The ``ramdisk`` deployment driver interface would be an opt-in interface"},{"line_number":55,"context_line":"which focuses on supporting ramdisk booting cases."},{"line_number":56,"context_line":""},{"line_number":57,"context_line":"The ``ramdisk`` deployment driver would also loop in the agent interface"},{"line_number":58,"context_line":"classes, in order to minimize complexity and allow for cleaning workflow"}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_1bcb91a4","line":55,"in_reply_to":"5f7c97a3_5bca7247","updated":"2018-07-24 14:17:00.000000000","message":"I had just assumed we\u0027d enable it wherever it made sense (not as the default though).","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"f413f80f4bb081ac95a684b667d7e0a6cf2904fc","unresolved":false,"context_lines":[{"line_number":66,"context_line":"  Effectively any image contents would be ignored."},{"line_number":67,"context_line":""},{"line_number":68,"context_line":"* The instance ``boot_option`` will be ignored by the"},{"line_number":69,"context_line":"  ramdisk interface, and will be explicitly set to ``ramdisk``."},{"line_number":70,"context_line":""},{"line_number":71,"context_line":"* The same ``kernel`` and ``ramdisk`` parameters that are used"},{"line_number":72,"context_line":"  for a \"glance\" image based deployment could be re-used to contain"}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_7bc72e3f","line":69,"updated":"2018-07-24 08:43:26.000000000","message":"++ to a warning","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":6618,"name":"Ruby Loo","email":"opensrloo@gmail.com","username":"rloo"},"change_message_id":"a173a646bc6803bd0f96af27053501923af9ecd0","unresolved":false,"context_lines":[{"line_number":66,"context_line":"  Effectively any image contents would be ignored."},{"line_number":67,"context_line":""},{"line_number":68,"context_line":"* The instance ``boot_option`` will be ignored by the"},{"line_number":69,"context_line":"  ramdisk interface, and will be explicitly set to ``ramdisk``."},{"line_number":70,"context_line":""},{"line_number":71,"context_line":"* The same ``kernel`` and ``ramdisk`` parameters that are used"},{"line_number":72,"context_line":"  for a \"glance\" image based deployment could be re-used to contain"}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_f2f3786c","line":69,"updated":"2018-07-23 14:38:06.000000000","message":"Will we log a warning (or return an error) if the user tries to set this value? (even if we don\u0027t mention it here, will need to know what to do in the code patch)","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":14826,"name":"Mark Goddard","email":"markgoddard86@gmail.com","username":"mgoddard"},"change_message_id":"59930fe3853fc7f6d8822687e7725abd466b5575","unresolved":false,"context_lines":[{"line_number":66,"context_line":"  Effectively any image contents would be ignored."},{"line_number":67,"context_line":""},{"line_number":68,"context_line":"* The instance ``boot_option`` will be ignored by the"},{"line_number":69,"context_line":"  ramdisk interface, and will be explicitly set to ``ramdisk``."},{"line_number":70,"context_line":""},{"line_number":71,"context_line":"* The same ``kernel`` and ``ramdisk`` parameters that are used"},{"line_number":72,"context_line":"  for a \"glance\" image based deployment could be re-used to contain"}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_068251c8","line":69,"in_reply_to":"5f7c97a3_7bc72e3f","updated":"2018-07-24 18:23:40.000000000","message":"Not sure I completely get how/why this needs to be different from netboot.","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"416bae064bacc2c4f3ab1bc30bec3efa02fcef71","unresolved":false,"context_lines":[{"line_number":66,"context_line":"  Effectively any image contents would be ignored."},{"line_number":67,"context_line":""},{"line_number":68,"context_line":"* The instance ``boot_option`` will be ignored by the"},{"line_number":69,"context_line":"  ramdisk interface, and will be explicitly set to ``ramdisk``."},{"line_number":70,"context_line":""},{"line_number":71,"context_line":"* The same ``kernel`` and ``ramdisk`` parameters that are used"},{"line_number":72,"context_line":"  for a \"glance\" image based deployment could be re-used to contain"}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_dd1a0dc6","line":69,"in_reply_to":"5f7c97a3_f2f3786c","updated":"2018-07-23 15:01:24.000000000","message":"I _think_ i put in a warning message into the code if overriden, I\u0027ll have to check.  Warning seems reasonable to me.","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":6618,"name":"Ruby Loo","email":"opensrloo@gmail.com","username":"rloo"},"change_message_id":"a173a646bc6803bd0f96af27053501923af9ecd0","unresolved":false,"context_lines":[{"line_number":85,"context_line":""},{"line_number":86,"context_line":"* Users must explicitly allow TFTP and/or HTTP iPXE endpoint access for"},{"line_number":87,"context_line":"  booting the baremetal nodes. Typically deployments do not allow tenant"},{"line_number":88,"context_line":"  networks are not permitted access to these endpoints."},{"line_number":89,"context_line":""},{"line_number":90,"context_line":"* Configuration drives would be ignored by ironic, and users would be"},{"line_number":91,"context_line":"  advised to make use of the metadata service."}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_527d8c1c","line":88,"updated":"2018-07-23 14:38:06.000000000","message":"There is something grammatically odd about this sentence.","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"416bae064bacc2c4f3ab1bc30bec3efa02fcef71","unresolved":false,"context_lines":[{"line_number":85,"context_line":""},{"line_number":86,"context_line":"* Users must explicitly allow TFTP and/or HTTP iPXE endpoint access for"},{"line_number":87,"context_line":"  booting the baremetal nodes. Typically deployments do not allow tenant"},{"line_number":88,"context_line":"  networks are not permitted access to these endpoints."},{"line_number":89,"context_line":""},{"line_number":90,"context_line":"* Configuration drives would be ignored by ironic, and users would be"},{"line_number":91,"context_line":"  advised to make use of the metadata service."}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_3d630148","line":88,"in_reply_to":"5f7c97a3_527d8c1c","updated":"2018-07-23 15:01:24.000000000","message":"s/are/to be/    s/access/to access/","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":6618,"name":"Ruby Loo","email":"opensrloo@gmail.com","username":"rloo"},"change_message_id":"a173a646bc6803bd0f96af27053501923af9ecd0","unresolved":false,"context_lines":[{"line_number":88,"context_line":"  networks are not permitted access to these endpoints."},{"line_number":89,"context_line":""},{"line_number":90,"context_line":"* Configuration drives would be ignored by ironic, and users would be"},{"line_number":91,"context_line":"  advised to make use of the metadata service."},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"  * Operators who directly deploy nodes using ironic may need to pass"},{"line_number":94,"context_line":"    additional kernel command line arguments to the node being \"deployed\""}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_d28bdcbe","line":91,"updated":"2018-07-23 14:38:06.000000000","message":"ditto about logging warning or emitting error, vs \u0027ignoring\u0027. Config drive is specified in the API request to deploy or rebuild, so I think we should just return an error, vs ignoring.","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"f413f80f4bb081ac95a684b667d7e0a6cf2904fc","unresolved":false,"context_lines":[{"line_number":88,"context_line":"  networks are not permitted access to these endpoints."},{"line_number":89,"context_line":""},{"line_number":90,"context_line":"* Configuration drives would be ignored by ironic, and users would be"},{"line_number":91,"context_line":"  advised to make use of the metadata service."},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"  * Operators who directly deploy nodes using ironic may need to pass"},{"line_number":94,"context_line":"    additional kernel command line arguments to the node being \"deployed\""}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_1bd4fae9","line":91,"updated":"2018-07-24 08:43:26.000000000","message":"yeah, a lot of deployments tell nova to always generate configdrives, but a warning is useful","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"a518b1f630bbc13f8b19b30725aa7cd4d93ae415","unresolved":false,"context_lines":[{"line_number":88,"context_line":"  networks are not permitted access to these endpoints."},{"line_number":89,"context_line":""},{"line_number":90,"context_line":"* Configuration drives would be ignored by ironic, and users would be"},{"line_number":91,"context_line":"  advised to make use of the metadata service."},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"  * Operators who directly deploy nodes using ironic may need to pass"},{"line_number":94,"context_line":"    additional kernel command line arguments to the node being \"deployed\""}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_67d22c36","line":91,"in_reply_to":"5f7c97a3_1bd4fae9","updated":"2018-07-24 11:59:18.000000000","message":"I agree, they have asked for it in the past and the community pushed back. We should go ahead and just do that as a minor RFE for driver level argument passing...  Realistically it might even be less than 10 lines of code....\n\nI\u0027m pushing for separate because it has out of scope implications and becomes scope creep for this single feature.","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":6618,"name":"Ruby Loo","email":"opensrloo@gmail.com","username":"rloo"},"change_message_id":"d3667be04c0c90fb39e01bbe1f9cea67f5adcdcc","unresolved":false,"context_lines":[{"line_number":88,"context_line":"  networks are not permitted access to these endpoints."},{"line_number":89,"context_line":""},{"line_number":90,"context_line":"* Configuration drives would be ignored by ironic, and users would be"},{"line_number":91,"context_line":"  advised to make use of the metadata service."},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"  * Operators who directly deploy nodes using ironic may need to pass"},{"line_number":94,"context_line":"    additional kernel command line arguments to the node being \"deployed\""}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_9b4b2127","line":91,"in_reply_to":"5f7c97a3_67d22c36","updated":"2018-07-24 14:17:00.000000000","message":"Ok, warning then. We just can\u0027t \u0027ignore\u0027 :)\n\nI\u0027m worried about non-nova deployments, where someone might say, \u0027hey, i specified it in the REST API, why didn\u0027t you just error instead of deploying as if everything was OK.\u0027","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"533a30ee96947df76bf40012ea7e6ecf8f681fab","unresolved":false,"context_lines":[{"line_number":88,"context_line":"  networks are not permitted access to these endpoints."},{"line_number":89,"context_line":""},{"line_number":90,"context_line":"* Configuration drives would be ignored by ironic, and users would be"},{"line_number":91,"context_line":"  advised to make use of the metadata service."},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"  * Operators who directly deploy nodes using ironic may need to pass"},{"line_number":94,"context_line":"    additional kernel command line arguments to the node being \"deployed\""}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_5b636046","line":91,"in_reply_to":"5f7c97a3_7d5ad982","updated":"2018-07-23 19:23:58.000000000","message":"Though about this, we should just log a warning, otherwise this would be completely incompatible with nova operating modes (i.e. whenever not a pure metadata service deployment)","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"416bae064bacc2c4f3ab1bc30bec3efa02fcef71","unresolved":false,"context_lines":[{"line_number":88,"context_line":"  networks are not permitted access to these endpoints."},{"line_number":89,"context_line":""},{"line_number":90,"context_line":"* Configuration drives would be ignored by ironic, and users would be"},{"line_number":91,"context_line":"  advised to make use of the metadata service."},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"  * Operators who directly deploy nodes using ironic may need to pass"},{"line_number":94,"context_line":"    additional kernel command line arguments to the node being \"deployed\""}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_7d5ad982","line":91,"in_reply_to":"5f7c97a3_d28bdcbe","updated":"2018-07-23 15:01:24.000000000","message":"I guess it makes sense to raise an exception. Likley best as a follow-up since that would end up being code in the conductor (which would keep that patch kind of small).","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"f413f80f4bb081ac95a684b667d7e0a6cf2904fc","unresolved":false,"context_lines":[{"line_number":92,"context_line":""},{"line_number":93,"context_line":"  * Operators who directly deploy nodes using ironic may need to pass"},{"line_number":94,"context_line":"    additional kernel command line arguments to the node being \"deployed\""},{"line_number":95,"context_line":"    via a ramdisk. In this case, a ``/instance_info/ramdisk_kernel_arguments``"},{"line_number":96,"context_line":"    field will be accepted to allow those operators to pass information to the"},{"line_number":97,"context_line":"    instance."},{"line_number":98,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_3bd1b6f9","line":95,"updated":"2018-07-24 08:43:26.000000000","message":"I still wonder if it could be useful in driver_info as well.. after all, the operators may want to add something permanent there. like, dunno, enable ibft. we can follow-up with it if we find it useful.","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":6618,"name":"Ruby Loo","email":"opensrloo@gmail.com","username":"rloo"},"change_message_id":"d3667be04c0c90fb39e01bbe1f9cea67f5adcdcc","unresolved":false,"context_lines":[{"line_number":92,"context_line":""},{"line_number":93,"context_line":"  * Operators who directly deploy nodes using ironic may need to pass"},{"line_number":94,"context_line":"    additional kernel command line arguments to the node being \"deployed\""},{"line_number":95,"context_line":"    via a ramdisk. In this case, a ``/instance_info/ramdisk_kernel_arguments``"},{"line_number":96,"context_line":"    field will be accepted to allow those operators to pass information to the"},{"line_number":97,"context_line":"    instance."},{"line_number":98,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_fb0815d9","line":95,"in_reply_to":"5f7c97a3_3bd1b6f9","updated":"2018-07-24 14:17:00.000000000","message":"Oh, good point. and yes, follow-up :)\n\nEven if we add to driver_info, that info will end up in node.instance_info I think, prior to deployment :)","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":6618,"name":"Ruby Loo","email":"opensrloo@gmail.com","username":"rloo"},"change_message_id":"a173a646bc6803bd0f96af27053501923af9ecd0","unresolved":false,"context_lines":[{"line_number":103,"context_line":"be in the underlying ``pxe`` boot interface, as the ramdisk interface"},{"line_number":104,"context_line":"overlays the ``pxe`` interface. Support for Ceph RBD or NFS based"},{"line_number":105,"context_line":"booting, from ``volume target`` information, would be a future"},{"line_number":106,"context_line":"enhanement."},{"line_number":107,"context_line":""},{"line_number":108,"context_line":"Alternatives"},{"line_number":109,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_92c004fc","line":106,"updated":"2018-07-23 14:38:06.000000000","message":"nit s/enhanement/enhancement/","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":6618,"name":"Ruby Loo","email":"opensrloo@gmail.com","username":"rloo"},"change_message_id":"a173a646bc6803bd0f96af27053501923af9ecd0","unresolved":false,"context_lines":[{"line_number":233,"context_line":"Upgrades and Backwards Compatibility"},{"line_number":234,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":235,"context_line":""},{"line_number":236,"context_line":"None"},{"line_number":237,"context_line":""},{"line_number":238,"context_line":"Documentation Impact"},{"line_number":239,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_52f90c80","line":236,"updated":"2018-07-23 14:38:06.000000000","message":"We haven\u0027t ever pinned new hw types or interfaces with Bare Metal API versions, right?\n\nI guess our assumption would be that no one would use any new stuff until *after* a rolling upgrade is complete, even though we don\u0027t guard against it.","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"416bae064bacc2c4f3ab1bc30bec3efa02fcef71","unresolved":false,"context_lines":[{"line_number":233,"context_line":"Upgrades and Backwards Compatibility"},{"line_number":234,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":235,"context_line":""},{"line_number":236,"context_line":"None"},{"line_number":237,"context_line":""},{"line_number":238,"context_line":"Documentation Impact"},{"line_number":239,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":11,"id":"5f7c97a3_9de9f583","line":236,"in_reply_to":"5f7c97a3_52f90c80","updated":"2018-07-23 15:01:24.000000000","message":"We have not because then we would run into difficulty with third party drivers and we would have to be aware of them to truly enforce it. Since we\u0027re not, I think it is best to avoid and operate on the premise that the API is and shall be agnostic to the actual field contents.","commit_id":"7693c94f2bd89c07e5853fc46e24361173de2294"}]}