)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":7730,"name":"Sahid Orentino Ferdjaoui","email":"sahid.ferdjaoui@industrialdiscipline.com","username":"sahid"},"change_message_id":"4c3779b624730586ebe2500efbcbaeb55f959549","unresolved":false,"context_lines":[{"line_number":5,"context_line":"CommitDate: 2018-06-01 13:40:38 +0000"},{"line_number":6,"context_line":""},{"line_number":7,"context_line":"nova improvement of maximum attach volumes more than 26 vols"},{"line_number":8,"context_line":""},{"line_number":9,"context_line":"Change-Id: I227b11d9ec0c2aebaeb0366fd731f829f21f1ca2"},{"line_number":10,"context_line":"Implements: blueprint nova-improvement-of-maximum-attach-volumes-more-than-26-vols"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":17,"id":"5f7c97a3_a9a29cac","line":8,"updated":"2018-06-01 13:46:40.000000000","message":"Can you fill the commit message with a bit more of information?","commit_id":"d2f151b311a8e95a8833e37f517a0ae83225d7a1"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"ec335164d18bd541abdaa520807008e53e0faac1","unresolved":false,"context_lines":[{"line_number":5,"context_line":"CommitDate: 2018-06-01 13:40:38 +0000"},{"line_number":6,"context_line":""},{"line_number":7,"context_line":"nova improvement of maximum attach volumes more than 26 vols"},{"line_number":8,"context_line":""},{"line_number":9,"context_line":"Change-Id: I227b11d9ec0c2aebaeb0366fd731f829f21f1ca2"},{"line_number":10,"context_line":"Implements: blueprint nova-improvement-of-maximum-attach-volumes-more-than-26-vols"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":17,"id":"5f7c97a3_ff5b661e","line":8,"in_reply_to":"5f7c97a3_a9a29cac","updated":"2018-06-01 15:00:59.000000000","message":"Done","commit_id":"d2f151b311a8e95a8833e37f517a0ae83225d7a1"},{"author":{"_account_id":7730,"name":"Sahid Orentino Ferdjaoui","email":"sahid.ferdjaoui@industrialdiscipline.com","username":"sahid"},"change_message_id":"63dce05eb4be0b9e8e2d4befea3cc9b05e98520c","unresolved":false,"context_lines":[{"line_number":7,"context_line":"nova improvement of maximum attach volumes more than 26 vols"},{"line_number":8,"context_line":""},{"line_number":9,"context_line":"Change-Id: I227b11d9ec0c2aebaeb0366fd731f829f21f1ca2"},{"line_number":10,"context_line":"Implements: blueprint nova-improvement-of-maximum-attach-volumes-more-than-26-vols"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":18,"id":"5f7c97a3_c11eb5ba","line":10,"updated":"2018-06-04 09:21:37.000000000","message":"I would have considered that as bug fix. The bug is the impossibility of blk_device_name() to handle more device names than 26 where virtio-scsi could handle much more.","commit_id":"2ee20418f410ba257f005a2406e6e51a1a55f515"},{"author":{"_account_id":7730,"name":"Sahid Orentino Ferdjaoui","email":"sahid.ferdjaoui@industrialdiscipline.com","username":"sahid"},"change_message_id":"1d165964fa6f4fc5851bcdd7ad59732c83a7d3f8","unresolved":false,"context_lines":[{"line_number":7,"context_line":"nova improvement of maximum attach volumes more than 26 vols"},{"line_number":8,"context_line":""},{"line_number":9,"context_line":"Change-Id: I227b11d9ec0c2aebaeb0366fd731f829f21f1ca2"},{"line_number":10,"context_line":"Implements: blueprint nova-improvement-of-maximum-attach-volumes-more-than-26-vols"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":18,"id":"5f7c97a3_d95e9775","line":10,"in_reply_to":"5f7c97a3_82daed48","updated":"2018-06-05 07:56:14.000000000","message":"Oh sorry Tsuyoshi it seems that I gave you some bad indications :/","commit_id":"2ee20418f410ba257f005a2406e6e51a1a55f515"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"c2c4086643bc569d3a834a15efa0f42c70b44927","unresolved":false,"context_lines":[{"line_number":7,"context_line":"nova improvement of maximum attach volumes more than 26 vols"},{"line_number":8,"context_line":""},{"line_number":9,"context_line":"Change-Id: I227b11d9ec0c2aebaeb0366fd731f829f21f1ca2"},{"line_number":10,"context_line":"Implements: blueprint nova-improvement-of-maximum-attach-volumes-more-than-26-vols"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":18,"id":"5f7c97a3_82daed48","line":10,"in_reply_to":"5f7c97a3_83867c13","updated":"2018-06-05 00:23:34.000000000","message":"Done","commit_id":"2ee20418f410ba257f005a2406e6e51a1a55f515"},{"author":{"_account_id":4690,"name":"melanie witt","display_name":"melwitt","email":"melwittt@gmail.com","username":"melwitt"},"change_message_id":"842bdaf27106934153f1dd4c1bf314e2a59ada0e","unresolved":false,"context_lines":[{"line_number":7,"context_line":"nova improvement of maximum attach volumes more than 26 vols"},{"line_number":8,"context_line":""},{"line_number":9,"context_line":"Change-Id: I227b11d9ec0c2aebaeb0366fd731f829f21f1ca2"},{"line_number":10,"context_line":"Implements: blueprint nova-improvement-of-maximum-attach-volumes-more-than-26-vols"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":18,"id":"5f7c97a3_83867c13","line":10,"in_reply_to":"5f7c97a3_c11eb5ba","updated":"2018-06-04 14:56:35.000000000","message":"FWIW, I think this is better represented as a feature in this case because originally the number of volumes was limited at 26 because of a limitation in libvirt at the time (if I understand correctly). By now, things have changed and we can increase the limit, but we can do so as a result of a lower-level change and can propagate that to users.","commit_id":"2ee20418f410ba257f005a2406e6e51a1a55f515"}],"nova/tests/unit/virt/libvirt/test_blockinfo.py":[{"author":{"_account_id":4690,"name":"melanie witt","display_name":"melwitt","email":"melwittt@gmail.com","username":"melwitt"},"change_message_id":"5b63d28149abc80fd656638eb2d9441c48018dc6","unresolved":false,"context_lines":[{"line_number":156,"context_line":"            \u0027/dev/vdw\u0027: {\u0027dev\u0027: \u0027vdw\u0027, \u0027bus\u0027: \u0027virtio\u0027, \u0027type\u0027: \u0027disk\u0027},"},{"line_number":157,"context_line":"            \u0027/dev/vdx\u0027: {\u0027dev\u0027: \u0027vdx\u0027, \u0027bus\u0027: \u0027virtio\u0027, \u0027type\u0027: \u0027disk\u0027},"},{"line_number":158,"context_line":"            \u0027/dev/vdy\u0027: {\u0027dev\u0027: \u0027vdy\u0027, \u0027bus\u0027: \u0027virtio\u0027, \u0027type\u0027: \u0027disk\u0027},"},{"line_number":159,"context_line":"            \u0027/dev/vdz\u0027: {\u0027dev\u0027: \u0027vdz\u0027, \u0027bus\u0027: \u0027virtio\u0027, \u0027type\u0027: \u0027disk\u0027},"},{"line_number":160,"context_line":"            }"},{"line_number":161,"context_line":"        dev \u003d blockinfo.find_disk_dev_for_disk_bus(mapping26vols, \u0027virtio\u0027)"},{"line_number":162,"context_line":"        self.assertEqual(\u0027vdaa\u0027, dev)"}],"source_content_type":"text/x-python","patch_set":11,"id":"5f7c97a3_1d3de52b","line":159,"updated":"2018-05-17 16:48:04.000000000","message":"For this you could have less lines of code by doing something like:\n\n mapping26vols \u003d {}\n for n in range(ord(\u0027a\u0027), ord(\u0027z\u0027)):\n     name \u003d \u0027vd%s\u0027 % chr(n)\n     mapping26vols[\u0027/dev/%s\u0027 % name] \u003d {\u0027dev\u0027: name, \u0027bus\u0027: \u0027virtio\u0027, \u0027type\u0027: \u0027disk\u0027}","commit_id":"576f360ac9318e230fc31e58f88532c8edaf461d"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"de1b94d861d2cf6721695da90f37f1cfd146c2e9","unresolved":false,"context_lines":[{"line_number":156,"context_line":"            \u0027/dev/vdw\u0027: {\u0027dev\u0027: \u0027vdw\u0027, \u0027bus\u0027: \u0027virtio\u0027, \u0027type\u0027: \u0027disk\u0027},"},{"line_number":157,"context_line":"            \u0027/dev/vdx\u0027: {\u0027dev\u0027: \u0027vdx\u0027, \u0027bus\u0027: \u0027virtio\u0027, \u0027type\u0027: \u0027disk\u0027},"},{"line_number":158,"context_line":"            \u0027/dev/vdy\u0027: {\u0027dev\u0027: \u0027vdy\u0027, \u0027bus\u0027: \u0027virtio\u0027, \u0027type\u0027: \u0027disk\u0027},"},{"line_number":159,"context_line":"            \u0027/dev/vdz\u0027: {\u0027dev\u0027: \u0027vdz\u0027, \u0027bus\u0027: \u0027virtio\u0027, \u0027type\u0027: \u0027disk\u0027},"},{"line_number":160,"context_line":"            }"},{"line_number":161,"context_line":"        dev \u003d blockinfo.find_disk_dev_for_disk_bus(mapping26vols, \u0027virtio\u0027)"},{"line_number":162,"context_line":"        self.assertEqual(\u0027vdaa\u0027, dev)"}],"source_content_type":"text/x-python","patch_set":11,"id":"5f7c97a3_62e22832","line":159,"in_reply_to":"5f7c97a3_1d3de52b","updated":"2018-05-18 03:55:33.000000000","message":"Done","commit_id":"576f360ac9318e230fc31e58f88532c8edaf461d"}],"nova/virt/libvirt/blockinfo.py":[{"author":{"_account_id":4690,"name":"melanie witt","display_name":"melwitt","email":"melwittt@gmail.com","username":"melwitt"},"change_message_id":"a2fa1e4111d4a96321746b2a072b36cc79aaf5af","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 1927634441"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":""},{"line_number":165,"context_line":"def blk_device_name(prefix, index):"}],"source_content_type":"text/x-python","patch_set":3,"id":"5f7c97a3_edec7790","line":162,"range":{"start_line":162,"start_character":15,"end_line":162,"end_character":25},"updated":"2018-05-11 01:22:09.000000000","message":"If we need to increase this value from 26, can we choose a more reasonable maximum number of disks for a single VM?","commit_id":"b0d9575ec0720a40bd446c4380e8105c89162a53"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"1fc04cc0d31816520c035d751ca510ac11a48722","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 1927634441"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":""},{"line_number":165,"context_line":"def blk_device_name(prefix, index):"}],"source_content_type":"text/x-python","patch_set":3,"id":"5f7c97a3_9c4f3633","line":162,"range":{"start_line":162,"start_character":15,"end_line":162,"end_character":25},"in_reply_to":"5f7c97a3_e802256f","updated":"2018-05-11 16:17:54.000000000","message":"Done","commit_id":"b0d9575ec0720a40bd446c4380e8105c89162a53"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"1372f9d796cb1123b02a7cdb04198c3385eaa42c","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 1927634441"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":""},{"line_number":165,"context_line":"def blk_device_name(prefix, index):"}],"source_content_type":"text/x-python","patch_set":3,"id":"5f7c97a3_e802256f","line":162,"range":{"start_line":162,"start_character":15,"end_line":162,"end_character":25},"in_reply_to":"5f7c97a3_edec7790","updated":"2018-05-11 02:24:36.000000000","message":"how about intmax?\n   INT_MAX:intmax /dev/vdfxshrxx 2147483647\n Its ok my SOC7 environment.","commit_id":"b0d9575ec0720a40bd446c4380e8105c89162a53"},{"author":{"_account_id":4690,"name":"melanie witt","display_name":"melwitt","email":"melwittt@gmail.com","username":"melwitt"},"change_message_id":"e11df37812b132d51942cbd0ff1b53ce888192bf","unresolved":false,"context_lines":[{"line_number":220,"context_line":"        \u0027uml\u0027: [\u0027uml\u0027],"},{"line_number":221,"context_line":"        \u0027lxc\u0027: [\u0027lxc\u0027],"},{"line_number":222,"context_line":"        \u0027parallels\u0027: [\u0027ide\u0027, \u0027scsi\u0027]"},{"line_number":223,"context_line":"    }"},{"line_number":224,"context_line":""},{"line_number":225,"context_line":"    if virt_type not in valid_bus:"},{"line_number":226,"context_line":"        raise exception.UnsupportedVirtType(virt\u003dvirt_type)"}],"source_content_type":"text/x-python","patch_set":3,"id":"5f7c97a3_4d1fabb6","line":223,"updated":"2018-05-11 00:53:27.000000000","message":"You have several formatting changes unrelated to the functionality for your patch. We want to avoid unrelated changes so that it is easier for reviewers to focus on the functional changes in the patch.","commit_id":"b0d9575ec0720a40bd446c4380e8105c89162a53"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"ba7a3633927ab59f63dc0c543940a25776abf387","unresolved":false,"context_lines":[{"line_number":220,"context_line":"        \u0027uml\u0027: [\u0027uml\u0027],"},{"line_number":221,"context_line":"        \u0027lxc\u0027: [\u0027lxc\u0027],"},{"line_number":222,"context_line":"        \u0027parallels\u0027: [\u0027ide\u0027, \u0027scsi\u0027]"},{"line_number":223,"context_line":"    }"},{"line_number":224,"context_line":""},{"line_number":225,"context_line":"    if virt_type not in valid_bus:"},{"line_number":226,"context_line":"        raise exception.UnsupportedVirtType(virt\u003dvirt_type)"}],"source_content_type":"text/x-python","patch_set":3,"id":"5f7c97a3_8fbf89cf","line":223,"in_reply_to":"5f7c97a3_4d1fabb6","updated":"2018-05-17 00:52:57.000000000","message":"Done","commit_id":"b0d9575ec0720a40bd446c4380e8105c89162a53"},{"author":{"_account_id":4690,"name":"melanie witt","display_name":"melwitt","email":"melwittt@gmail.com","username":"melwitt"},"change_message_id":"9a90ce15c5365d085cc8336933be99f659ca1442","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 2147483647"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":"def blk_device_name(prefix, index):"},{"line_number":165,"context_line":"    \"\"\"Returns device unit name by index"}],"source_content_type":"text/x-python","patch_set":5,"id":"5f7c97a3_7ce25aa2","line":162,"range":{"start_line":162,"start_character":15,"end_line":162,"end_character":25},"updated":"2018-05-11 04:14:41.000000000","message":"Sorry, what I meant earlier was, can we select a realistic upper limit for the number of disks connected to a single VM? Like 256? How many were you thinking for your use case?\n\nHaving a maximum of 2 billion doesn\u0027t make much sense, although I\u0027m guessing you were thinking this is just a good number as any if we\u0027re going to break the loop once we find an unassigned device.","commit_id":"085d458ac63660b7e5a05da8535e760b0f5f016e"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"de1a2b81f3f76efde9cd3eefdc7bb3a7de48aa78","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 2147483647"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":"def blk_device_name(prefix, index):"},{"line_number":165,"context_line":"    \"\"\"Returns device unit name by index"}],"source_content_type":"text/x-python","patch_set":5,"id":"5f7c97a3_6b203cad","line":162,"range":{"start_line":162,"start_character":15,"end_line":162,"end_character":25},"in_reply_to":"5f7c97a3_5ece65b4","updated":"2018-05-11 23:29:43.000000000","message":"Done","commit_id":"085d458ac63660b7e5a05da8535e760b0f5f016e"},{"author":{"_account_id":6962,"name":"Kashyap Chamarthy","email":"kchamart@redhat.com","username":"kashyapc"},"change_message_id":"365c20598495d32325315ac86004edf87acd03e4","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 2147483647"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":"def blk_device_name(prefix, index):"},{"line_number":165,"context_line":"    \"\"\"Returns device unit name by index"}],"source_content_type":"text/x-python","patch_set":5,"id":"5f7c97a3_f3ae0c97","line":162,"range":{"start_line":162,"start_character":15,"end_line":162,"end_character":25},"in_reply_to":"5f7c97a3_5ece65b4","updated":"2018-06-08 08:20:45.000000000","message":"Just catching up with this review; FWIW, I agree with Mel\u0027s comment that 1024 is a reasonable default in Nova\u0027s case.","commit_id":"085d458ac63660b7e5a05da8535e760b0f5f016e"},{"author":{"_account_id":6062,"name":"jichenjc","email":"jichenjc@cn.ibm.com","username":"jichenjc"},"change_message_id":"a318e2638bceec6d01a5356fbd6c7b232e2ef1a5","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 2147483647"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":"def blk_device_name(prefix, index):"},{"line_number":165,"context_line":"    \"\"\"Returns device unit name by index"}],"source_content_type":"text/x-python","patch_set":5,"id":"5f7c97a3_ab40a2a6","line":162,"range":{"start_line":162,"start_character":15,"end_line":162,"end_character":25},"in_reply_to":"5f7c97a3_7ce25aa2","updated":"2018-05-11 09:48:22.000000000","message":"googled and find this:\n\nhttps://elixir.bootlin.com/linux/v4.4/source/drivers/scsi/sd.c#L547\n\nhttps://www.reddit.com/r/linuxquestions/comments/4cifq3/max_number_of_devices_in_linux/\n\nso seems we can follow that ?","commit_id":"085d458ac63660b7e5a05da8535e760b0f5f016e"},{"author":{"_account_id":4690,"name":"melanie witt","display_name":"melwitt","email":"melwittt@gmail.com","username":"melwitt"},"change_message_id":"2af9c913a5f41d0b97acf06f22af1bcbbb0b9681","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 2147483647"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":"def blk_device_name(prefix, index):"},{"line_number":165,"context_line":"    \"\"\"Returns device unit name by index"}],"source_content_type":"text/x-python","patch_set":5,"id":"5f7c97a3_5ece65b4","line":162,"range":{"start_line":162,"start_character":15,"end_line":162,"end_character":25},"in_reply_to":"5f7c97a3_98f65563","updated":"2018-05-11 18:12:44.000000000","message":"Thanks for the detailed data. I\u0027ve been looking for answers about the practical limit too and a colleague shared this blog post [1] in which someone actually tested attaching 10,000 devices to a single VM.\n\nGiven that, I think we should just choose a value that is low/practical and satisfies the use case to allow attaching \u003e 26 volumes.\n\nI was thinking no larger than 1024, for example.\n\n[1] https://rwmj.wordpress.com/2017/04/25/how-many-disks-can-you-add-to-a-virtual-linux-machine/","commit_id":"085d458ac63660b7e5a05da8535e760b0f5f016e"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"d08ddc7541feb5c1c12b656c77f150da778559ec","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 2147483647"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":"def blk_device_name(prefix, index):"},{"line_number":165,"context_line":"    \"\"\"Returns device unit name by index"}],"source_content_type":"text/x-python","patch_set":5,"id":"5f7c97a3_98f65563","line":162,"range":{"start_line":162,"start_character":15,"end_line":162,"end_character":25},"in_reply_to":"5f7c97a3_ab40a2a6","updated":"2018-05-11 16:53:22.000000000","message":"vd the virtid_blk device assigned by pci-bus.\n  pci addressing was below\n\n(complex:32):(bus-no:256):(id:32)\n\neach pci-bus have 32 devices.\nmax pci-bus was 256.\ncomplex was matched by cpu socket.\ncurrently, intel\u0027s pci-complex was built by QPI-connection.\nQPI-CPU-connection is now 32.\nthen current maximum virtio_blk no is,\n\n  32 x 256 x 32 \u003d 262144\n\n\n    if disk_bus \u003d\u003d \"ide\":\n        return 4\n    elif disk_bus \u003d\u003d \"sd\":\n        return 262144\n    elif disk_bus \u003d\u003d \"vd\":\n        return 262144\n    else:\n        return 26\n\nIn old described information, maximum no of virtio_blk device was 32.(bus no\u003d0)\n(max scsi-id was 16 too!!)\nNow, qemu/kvm can handle multiple pci bus and increase max no of devices.","commit_id":"085d458ac63660b7e5a05da8535e760b0f5f016e"},{"author":{"_account_id":6962,"name":"Kashyap Chamarthy","email":"kchamart@redhat.com","username":"kashyapc"},"change_message_id":"074fa370e548688390849d206054024093b1f182","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 2147483647"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":"def blk_device_name(prefix, index):"},{"line_number":165,"context_line":"    \"\"\"Returns device unit name by index"}],"source_content_type":"text/x-python","patch_set":5,"id":"5f7c97a3_5bf7845a","line":162,"range":{"start_line":162,"start_character":15,"end_line":162,"end_character":25},"in_reply_to":"5f7c97a3_f3ae0c97","updated":"2018-06-08 08:23:18.000000000","message":"\u003e Just catching up with this review; FWIW, I agree with Mel\u0027s comment\n \u003e that 1024 is a reasonable default in Nova\u0027s case.\n\nOops, disregard the above stale comment on PS5.  I was supposed to remove it, but forgot to do so.","commit_id":"085d458ac63660b7e5a05da8535e760b0f5f016e"},{"author":{"_account_id":4690,"name":"melanie witt","display_name":"melwitt","email":"melwittt@gmail.com","username":"melwitt"},"change_message_id":"9a90ce15c5365d085cc8336933be99f659ca1442","unresolved":false,"context_lines":[{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 2147483647"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":"def blk_device_name(prefix, index):"},{"line_number":165,"context_line":"    \"\"\"Returns device unit name by index"},{"line_number":166,"context_line":"        i.e."}],"source_content_type":"text/x-python","patch_set":5,"id":"5f7c97a3_bcdc72e4","line":163,"updated":"2018-05-11 04:14:41.000000000","message":"There\u0027s a pep8 failure here -- must have two blank lines between globals.","commit_id":"085d458ac63660b7e5a05da8535e760b0f5f016e"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"de1a2b81f3f76efde9cd3eefdc7bb3a7de48aa78","unresolved":false,"context_lines":[{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 2147483647"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":"def blk_device_name(prefix, index):"},{"line_number":165,"context_line":"    \"\"\"Returns device unit name by index"},{"line_number":166,"context_line":"        i.e."}],"source_content_type":"text/x-python","patch_set":5,"id":"5f7c97a3_4b1b0080","line":163,"in_reply_to":"5f7c97a3_bcdc72e4","updated":"2018-05-11 23:29:43.000000000","message":"Done","commit_id":"085d458ac63660b7e5a05da8535e760b0f5f016e"},{"author":{"_account_id":6062,"name":"jichenjc","email":"jichenjc@cn.ibm.com","username":"jichenjc"},"change_message_id":"a318e2638bceec6d01a5356fbd6c7b232e2ef1a5","unresolved":false,"context_lines":[{"line_number":165,"context_line":"    \"\"\"Returns device unit name by index"},{"line_number":166,"context_line":"        i.e."},{"line_number":167,"context_line":"            vda, vdb,..., vdzzz"},{"line_number":168,"context_line":"    \"\"\""},{"line_number":169,"context_line":"    base \u003d ord(\u0027z\u0027) - ord(\u0027a\u0027) + 1"},{"line_number":170,"context_line":"    unitdevname \u003d \"\""},{"line_number":171,"context_line":"    while True:"},{"line_number":172,"context_line":"        p \u003d chr(ord(\u0027a\u0027) + (index % base))"},{"line_number":173,"context_line":"        unitdevname \u003d p + unitdevname"},{"line_number":174,"context_line":"        index \u003d (index / base) - 1"},{"line_number":175,"context_line":"        if (index \u003c 0):"},{"line_number":176,"context_line":"            break"},{"line_number":177,"context_line":""},{"line_number":178,"context_line":"    return prefix + unitdevname"}],"source_content_type":"text/x-python","patch_set":5,"id":"5f7c97a3_cb3dde2b","line":175,"range":{"start_line":168,"start_character":4,"end_line":175,"end_character":23},"updated":"2018-05-11 09:48:22.000000000","message":"ut is needed","commit_id":"085d458ac63660b7e5a05da8535e760b0f5f016e"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"c3122e969a84b03ee4044b17813db9bd629b537b","unresolved":false,"context_lines":[{"line_number":165,"context_line":"    \"\"\"Returns device unit name by index"},{"line_number":166,"context_line":"        i.e."},{"line_number":167,"context_line":"            vda, vdb,..., vdzzz"},{"line_number":168,"context_line":"    \"\"\""},{"line_number":169,"context_line":"    base \u003d ord(\u0027z\u0027) - ord(\u0027a\u0027) + 1"},{"line_number":170,"context_line":"    unitdevname \u003d \"\""},{"line_number":171,"context_line":"    while True:"},{"line_number":172,"context_line":"        p \u003d chr(ord(\u0027a\u0027) + (index % base))"},{"line_number":173,"context_line":"        unitdevname \u003d p + unitdevname"},{"line_number":174,"context_line":"        index \u003d (index / base) - 1"},{"line_number":175,"context_line":"        if (index \u003c 0):"},{"line_number":176,"context_line":"            break"},{"line_number":177,"context_line":""},{"line_number":178,"context_line":"    return prefix + unitdevname"}],"source_content_type":"text/x-python","patch_set":5,"id":"5f7c97a3_6fd1cd2f","line":175,"range":{"start_line":168,"start_character":4,"end_line":175,"end_character":23},"in_reply_to":"5f7c97a3_98a175fb","updated":"2018-05-17 00:48:07.000000000","message":"Done","commit_id":"085d458ac63660b7e5a05da8535e760b0f5f016e"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"1fc04cc0d31816520c035d751ca510ac11a48722","unresolved":false,"context_lines":[{"line_number":165,"context_line":"    \"\"\"Returns device unit name by index"},{"line_number":166,"context_line":"        i.e."},{"line_number":167,"context_line":"            vda, vdb,..., vdzzz"},{"line_number":168,"context_line":"    \"\"\""},{"line_number":169,"context_line":"    base \u003d ord(\u0027z\u0027) - ord(\u0027a\u0027) + 1"},{"line_number":170,"context_line":"    unitdevname \u003d \"\""},{"line_number":171,"context_line":"    while True:"},{"line_number":172,"context_line":"        p \u003d chr(ord(\u0027a\u0027) + (index % base))"},{"line_number":173,"context_line":"        unitdevname \u003d p + unitdevname"},{"line_number":174,"context_line":"        index \u003d (index / base) - 1"},{"line_number":175,"context_line":"        if (index \u003c 0):"},{"line_number":176,"context_line":"            break"},{"line_number":177,"context_line":""},{"line_number":178,"context_line":"    return prefix + unitdevname"}],"source_content_type":"text/x-python","patch_set":5,"id":"5f7c97a3_98a175fb","line":175,"range":{"start_line":168,"start_character":4,"end_line":175,"end_character":23},"in_reply_to":"5f7c97a3_cb3dde2b","updated":"2018-05-11 16:17:54.000000000","message":"I made this code from latest linux virtio_blk code.\nhttps://github.com/torvalds/linux/blob/master/drivers/block/virtio_blk.c#L507\n\nI did this unit test.\nidx \u003d 0\nwhile (idx \u003c 200):\n print \"%d %s\" %( idx, blk_device_name(\"vd\", idx))\n idx \u003d idx +1\n\nthis get results\n\n0 vda\n1 vdb\n  :\n24 vdy\n25 vdz\n26 vdaa\n27 vdab\n28 vdac\n  :","commit_id":"085d458ac63660b7e5a05da8535e760b0f5f016e"},{"author":{"_account_id":4690,"name":"melanie witt","display_name":"melwitt","email":"melwittt@gmail.com","username":"melwitt"},"change_message_id":"9a90ce15c5365d085cc8336933be99f659ca1442","unresolved":false,"context_lines":[{"line_number":176,"context_line":"            break"},{"line_number":177,"context_line":""},{"line_number":178,"context_line":"    return prefix + unitdevname"},{"line_number":179,"context_line":""},{"line_number":180,"context_line":"def find_disk_dev_for_disk_bus(mapping, bus,"},{"line_number":181,"context_line":"                               assigned_devices\u003dNone):"},{"line_number":182,"context_line":"    \"\"\"Identify a free disk dev name for a bus."}],"source_content_type":"text/x-python","patch_set":5,"id":"5f7c97a3_1cd486bc","line":179,"updated":"2018-05-11 04:14:41.000000000","message":"There\u0027s a pep8 failure here -- must have two blank lines between globals.","commit_id":"085d458ac63660b7e5a05da8535e760b0f5f016e"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"de1a2b81f3f76efde9cd3eefdc7bb3a7de48aa78","unresolved":false,"context_lines":[{"line_number":176,"context_line":"            break"},{"line_number":177,"context_line":""},{"line_number":178,"context_line":"    return prefix + unitdevname"},{"line_number":179,"context_line":""},{"line_number":180,"context_line":"def find_disk_dev_for_disk_bus(mapping, bus,"},{"line_number":181,"context_line":"                               assigned_devices\u003dNone):"},{"line_number":182,"context_line":"    \"\"\"Identify a free disk dev name for a bus."}],"source_content_type":"text/x-python","patch_set":5,"id":"5f7c97a3_0b11885d","line":179,"in_reply_to":"5f7c97a3_1cd486bc","updated":"2018-05-11 23:29:43.000000000","message":"Done","commit_id":"085d458ac63660b7e5a05da8535e760b0f5f016e"},{"author":{"_account_id":4690,"name":"melanie witt","display_name":"melwitt","email":"melwittt@gmail.com","username":"melwitt"},"change_message_id":"3e0aaa79978819fd74faa2c15771b5a11042507e","unresolved":false,"context_lines":[{"line_number":180,"context_line":"        if (index \u003c 0):"},{"line_number":181,"context_line":"            break"},{"line_number":182,"context_line":""},{"line_number":183,"context_line":"    return prefix + unitdevname"},{"line_number":184,"context_line":""},{"line_number":185,"context_line":""},{"line_number":186,"context_line":"def find_disk_dev_for_disk_bus(mapping, bus,"}],"source_content_type":"text/x-python","patch_set":6,"id":"5f7c97a3_010b84d5","line":183,"updated":"2018-05-15 01:52:08.000000000","message":"For this method, I would add a test called \"test_blk_device_name\" where you call it several times (maybe 5 or 7 times) with differnet specific prefix and index and then self.assert(expected, actual) where:\n\n actual \u003d blk_device_name(prefix, index)\n\nThis way, you will verify that the function returns the device names we expect. I would aim for examples names like \"vda\", \"vdaa\", \"vdabc\", \"vdzzz\" which are later in the name generation to verify that the passed index yields the correct expected pattern.","commit_id":"f0aebb5f2a5734b6adf40ec19e65cd837a263327"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"358350e214c2081441125f19a89172a3e74b9407","unresolved":false,"context_lines":[{"line_number":180,"context_line":"        if (index \u003c 0):"},{"line_number":181,"context_line":"            break"},{"line_number":182,"context_line":""},{"line_number":183,"context_line":"    return prefix + unitdevname"},{"line_number":184,"context_line":""},{"line_number":185,"context_line":""},{"line_number":186,"context_line":"def find_disk_dev_for_disk_bus(mapping, bus,"}],"source_content_type":"text/x-python","patch_set":6,"id":"5f7c97a3_e4bd1b02","line":183,"in_reply_to":"5f7c97a3_010b84d5","updated":"2018-05-16 02:02:32.000000000","message":"Done","commit_id":"f0aebb5f2a5734b6adf40ec19e65cd837a263327"},{"author":{"_account_id":4690,"name":"melanie witt","display_name":"melwitt","email":"melwittt@gmail.com","username":"melwitt"},"change_message_id":"3e0aaa79978819fd74faa2c15771b5a11042507e","unresolved":false,"context_lines":[{"line_number":210,"context_line":"        if not has_disk_dev(mapping, disk_dev):"},{"line_number":211,"context_line":"            if disk_dev not in assigned_devices:"},{"line_number":212,"context_line":"                return disk_dev"},{"line_number":213,"context_line":"        idx \u003d idx + 1"},{"line_number":214,"context_line":""},{"line_number":215,"context_line":"    msg \u003d _(\"No free disk device names for prefix \u0027%s\u0027\") % dev_prefix"},{"line_number":216,"context_line":"    raise exception.InternalError(msg)"}],"source_content_type":"text/x-python","patch_set":6,"id":"5f7c97a3_a1265067","line":213,"updated":"2018-05-15 01:52:08.000000000","message":"To test this, I would create a new test similar to \"test_find_disk_dev\" where you test some different disk_bus and verify the returned dev is as expected.","commit_id":"f0aebb5f2a5734b6adf40ec19e65cd837a263327"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"358350e214c2081441125f19a89172a3e74b9407","unresolved":false,"context_lines":[{"line_number":210,"context_line":"        if not has_disk_dev(mapping, disk_dev):"},{"line_number":211,"context_line":"            if disk_dev not in assigned_devices:"},{"line_number":212,"context_line":"                return disk_dev"},{"line_number":213,"context_line":"        idx \u003d idx + 1"},{"line_number":214,"context_line":""},{"line_number":215,"context_line":"    msg \u003d _(\"No free disk device names for prefix \u0027%s\u0027\") % dev_prefix"},{"line_number":216,"context_line":"    raise exception.InternalError(msg)"}],"source_content_type":"text/x-python","patch_set":6,"id":"5f7c97a3_a43963a4","line":213,"in_reply_to":"5f7c97a3_a1265067","updated":"2018-05-16 02:02:32.000000000","message":"Done","commit_id":"f0aebb5f2a5734b6adf40ec19e65cd837a263327"},{"author":{"_account_id":4690,"name":"melanie witt","display_name":"melwitt","email":"melwittt@gmail.com","username":"melwitt"},"change_message_id":"5b63d28149abc80fd656638eb2d9441c48018dc6","unresolved":false,"context_lines":[{"line_number":161,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":162,"context_line":"        return 1024"},{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"virtio\":"},{"line_number":164,"context_line":"        return 1024"},{"line_number":165,"context_line":"    else:"},{"line_number":166,"context_line":"        return 26"},{"line_number":167,"context_line":""}],"source_content_type":"text/x-python","patch_set":11,"id":"5f7c97a3_38722fe5","line":164,"updated":"2018-05-17 16:48:04.000000000","message":"I chatted with some other folks about this proposed maximum in #openstack-nova today and there\u0027s concern that 1024 is still too high to expect reasonable throughput from the devices attached to one instance. Please also check the new comments on the blueprint.\n\nCan you please describe your use case briefly? And tell us how many devices you need to attach?","commit_id":"576f360ac9318e230fc31e58f88532c8edaf461d"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"de1b94d861d2cf6721695da90f37f1cfd146c2e9","unresolved":false,"context_lines":[{"line_number":161,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":162,"context_line":"        return 1024"},{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"virtio\":"},{"line_number":164,"context_line":"        return 1024"},{"line_number":165,"context_line":"    else:"},{"line_number":166,"context_line":"        return 26"},{"line_number":167,"context_line":""}],"source_content_type":"text/x-python","patch_set":11,"id":"5f7c97a3_e281584c","line":164,"in_reply_to":"5f7c97a3_38722fe5","updated":"2018-05-18 03:55:33.000000000","message":"Done","commit_id":"576f360ac9318e230fc31e58f88532c8edaf461d"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"e26782c40b06b7cc393dfcb0d3e9a5bb07151cd6","unresolved":false,"context_lines":[{"line_number":161,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":162,"context_line":"        return 1024"},{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"virtio\":"},{"line_number":164,"context_line":"        return 1024"},{"line_number":165,"context_line":"    else:"},{"line_number":166,"context_line":"        return 26"},{"line_number":167,"context_line":""}],"source_content_type":"text/x-python","patch_set":11,"id":"5f7c97a3_d9d457c0","line":164,"in_reply_to":"5f7c97a3_38722fe5","updated":"2018-05-18 00:18:16.000000000","message":"Let\u0027s go haggling!!\nfor the MIDDLETEXT(mariadb limits), Please wait for evaluating on my environment(Suse Openstack Cloud 7:newton kvm). I\u0027ll try to 1024 and report on the blueprint.\n\nMy minimum goal is 64. Testing my hardware products with jenkins, openstack environment is match for CI/CD.","commit_id":"576f360ac9318e230fc31e58f88532c8edaf461d"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"de1b94d861d2cf6721695da90f37f1cfd146c2e9","unresolved":false,"context_lines":[{"line_number":161,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":162,"context_line":"        return 1024"},{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"virtio\":"},{"line_number":164,"context_line":"        return 1024"},{"line_number":165,"context_line":"    else:"},{"line_number":166,"context_line":"        return 26"},{"line_number":167,"context_line":""}],"source_content_type":"text/x-python","patch_set":11,"id":"5f7c97a3_82966494","line":164,"in_reply_to":"5f7c97a3_9c6fad91","updated":"2018-05-18 03:55:33.000000000","message":"My app is auto-testing ceph on openstack. a ceph-osd node has many volumes.\nif each instance can handle many volumes, I can solve provisioning SDS on openstack.\n(without buying real disks.)","commit_id":"576f360ac9318e230fc31e58f88532c8edaf461d"},{"author":{"_account_id":4690,"name":"melanie witt","display_name":"melwitt","email":"melwittt@gmail.com","username":"melwitt"},"change_message_id":"ef55e0b85a7e4ed8b6f37ab4b3743badc8cf66e9","unresolved":false,"context_lines":[{"line_number":161,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":162,"context_line":"        return 1024"},{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"virtio\":"},{"line_number":164,"context_line":"        return 1024"},{"line_number":165,"context_line":"    else:"},{"line_number":166,"context_line":"        return 26"},{"line_number":167,"context_line":""}],"source_content_type":"text/x-python","patch_set":11,"id":"5f7c97a3_9c6fad91","line":164,"in_reply_to":"5f7c97a3_d9d457c0","updated":"2018-05-18 01:00:06.000000000","message":"Heh. Well, really, we would just like to learn about your use case for this many volumes? What are you working on where you need a large number of volumes attach to one VM? We understand the \"what\" but we would like to understand the \"why\". Thanks.","commit_id":"576f360ac9318e230fc31e58f88532c8edaf461d"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"8cfa106619dd947b465819d7d79d200eaf8a536f","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 64"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":""},{"line_number":165,"context_line":"def blk_device_name(prefix, index):"}],"source_content_type":"text/x-python","patch_set":13,"id":"5f7c97a3_c1544857","line":162,"range":{"start_line":162,"start_character":0,"end_line":162,"end_character":17},"updated":"2018-05-18 22:59:24.000000000","message":"This code allways fails at Zuul.\n\nnova-cells-v1\tFAILURE in 50m 51s\n\nI\u0027ll back to code for kvm(virtio)\n\n+    elif disk_bus \u003d\u003d \"virtio\":\n+        return 256","commit_id":"10ebc2eedd644bc10bb4e7c6b70b5e93a3f6400c"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"babddfb719c31474babd3ccffb22a77f4f2a4d27","unresolved":false,"context_lines":[{"line_number":158,"context_line":""},{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    elif disk_bus \u003d\u003d \"virtio\":"},{"line_number":162,"context_line":"        return 256"},{"line_number":163,"context_line":"    else:"},{"line_number":164,"context_line":"        return 26"}],"source_content_type":"text/x-python","patch_set":14,"id":"5f7c97a3_b05e3743","line":161,"updated":"2018-05-19 03:09:59.000000000","message":"reverting for patch set 13","commit_id":"6e374899dbccaa9dd121b6e0c5118ed547061dd0"},{"author":{"_account_id":7730,"name":"Sahid Orentino Ferdjaoui","email":"sahid.ferdjaoui@industrialdiscipline.com","username":"sahid"},"change_message_id":"3bc0c8d21f684a50ea36b562c2a715512e1c699d","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 128"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":""},{"line_number":165,"context_line":"def blk_device_name(prefix, index):"}],"source_content_type":"text/x-python","patch_set":16,"id":"5f7c97a3_c696ecf0","line":162,"updated":"2018-06-01 08:08:46.000000000","message":"I don\u0027t think we can do that. We really have to add new conditions to handle the \u0027virtio-scsi\u0027 case.\n\n- virtio-blk is using for each disk, one PCI slot. Since there is a limit of 32 slots per machine (For Q35 it\u0027s different) and some of them are already used in our default guest configuration for networking devices, memory balloon device, USB controller...  My thinking is we should probably keep the limit of 26.\n\n- For virtio-scsi it\u0027s different, we have one controller that is using one PCI slot. Nova currently does not support to have more virtio-scsi controller. One controller can have 256 targets and each target supports 1 to 2^14 logical device (LUN). So it\u0027s about 4194304, which does not really make sense. Limiting it to 128 or perhaps the number of targets seems good.\n\n- For the native QEMU scsi implementation I think a target can only handle one LUN, meaning 256 devices.","commit_id":"3db7ebcd89d06554dc4650fdf3d6e5bf58c56e36"},{"author":{"_account_id":7730,"name":"Sahid Orentino Ferdjaoui","email":"sahid.ferdjaoui@industrialdiscipline.com","username":"sahid"},"change_message_id":"4c3779b624730586ebe2500efbcbaeb55f959549","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 128"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":""},{"line_number":165,"context_line":"def blk_device_name(prefix, index):"}],"source_content_type":"text/x-python","patch_set":16,"id":"5f7c97a3_c9c03853","line":162,"in_reply_to":"5f7c97a3_89cc80bb","updated":"2018-06-01 13:46:40.000000000","message":"Interesting point, for the virtio-blk you mentioned that, lastet linux and QEMU can handle more. \n\nWhat do you mean by \"lasted\" we have to ensure that the distributions are going to ship the good kernel and QEMU. What about to add a check based on versions?","commit_id":"3db7ebcd89d06554dc4650fdf3d6e5bf58c56e36"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"6b8bef11439d663752adb9f9663e08d0279a475c","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 128"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":""},{"line_number":165,"context_line":"def blk_device_name(prefix, index):"}],"source_content_type":"text/x-python","patch_set":16,"id":"5f7c97a3_89cc80bb","line":162,"in_reply_to":"5f7c97a3_c696ecf0","updated":"2018-06-01 13:32:15.000000000","message":"\u003e I don\u0027t think we can do that. We really have to add new conditions\n \u003e to handle the \u0027virtio-scsi\u0027 case.\n \u003e \n \u003e - virtio-blk is using for each disk, one PCI slot. Since there is a\n \u003e limit of 32 slots per machine (For Q35 it\u0027s different) and some of\n \u003e them are already used in our default guest configuration for\n \u003e networking devices, memory balloon device, USB controller...  My\n \u003e thinking is we should probably keep the limit of 26.\n \u003e \n\nno, latest linux and qemu can handle 32(slots) x 256(pcibus).\nvirtio-blk shoud be 8192(\u003d 32 x 256).\n\n \u003e - For virtio-scsi it\u0027s different, we have one controller that is\n \u003e using one PCI slot. Nova currently does not support to have more\n \u003e virtio-scsi controller. One controller can have 256 targets and\n \u003e each target supports 1 to 2^14 logical device (LUN). So it\u0027s about\n \u003e 4194304, which does not really make sense. Limiting it to 128 or\n \u003e perhaps the number of targets seems good.\n \u003e \n\nagree!\n\n \u003e - For the native QEMU scsi implementation I think a target can only\n \u003e handle one LUN, meaning 256 devices.\n\nagree!\n\nI think bus limit number shoud be specific limit, not a convinent number.\n\nI\u0027ll fix by \"specific based limitation\" on next patch set! thx!","commit_id":"3db7ebcd89d06554dc4650fdf3d6e5bf58c56e36"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"fd69205ea235448006f2b313b355eb2bdb707429","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 128"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":""},{"line_number":165,"context_line":"def blk_device_name(prefix, index):"}],"source_content_type":"text/x-python","patch_set":16,"id":"5f7c97a3_69b3240b","line":162,"in_reply_to":"5f7c97a3_c696ecf0","updated":"2018-06-01 13:42:48.000000000","message":"Done","commit_id":"3db7ebcd89d06554dc4650fdf3d6e5bf58c56e36"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"ec335164d18bd541abdaa520807008e53e0faac1","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    else:"},{"line_number":162,"context_line":"        return 128"},{"line_number":163,"context_line":""},{"line_number":164,"context_line":""},{"line_number":165,"context_line":"def blk_device_name(prefix, index):"}],"source_content_type":"text/x-python","patch_set":16,"id":"5f7c97a3_493bc836","line":162,"in_reply_to":"5f7c97a3_c9c03853","updated":"2018-06-01 15:00:59.000000000","message":"I made this code from *LATEST* linux virtio_blk code.\nSee \n https://github.com/torvalds/linux/blob/master/drivers/block/virtio_blk.c#L714","commit_id":"3db7ebcd89d06554dc4650fdf3d6e5bf58c56e36"},{"author":{"_account_id":7730,"name":"Sahid Orentino Ferdjaoui","email":"sahid.ferdjaoui@industrialdiscipline.com","username":"sahid"},"change_message_id":"4c3779b624730586ebe2500efbcbaeb55f959549","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    elif disk_bus \u003d\u003d \"virtio\":"},{"line_number":162,"context_line":"        return 4194304"},{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":164,"context_line":"        return 256"},{"line_number":165,"context_line":"    else:"}],"source_content_type":"text/x-python","patch_set":17,"id":"5f7c97a3_a9905c44","line":162,"updated":"2018-06-01 13:46:40.000000000","message":"You can\u0027t really do that, based on you unittest the max we handle is sdzzz which is 18277. That number does not really make sense so I\u0027m OK to see a reasonable limit.\n\nAlso I would have seen a comment explaining the numbers.","commit_id":"d2f151b311a8e95a8833e37f517a0ae83225d7a1"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"ec335164d18bd541abdaa520807008e53e0faac1","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    elif disk_bus \u003d\u003d \"virtio\":"},{"line_number":162,"context_line":"        return 4194304"},{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":164,"context_line":"        return 256"},{"line_number":165,"context_line":"    else:"}],"source_content_type":"text/x-python","patch_set":17,"id":"5f7c97a3_5fa29afd","line":162,"in_reply_to":"5f7c97a3_a9905c44","updated":"2018-06-01 15:00:59.000000000","message":"Done","commit_id":"d2f151b311a8e95a8833e37f517a0ae83225d7a1"},{"author":{"_account_id":7730,"name":"Sahid Orentino Ferdjaoui","email":"sahid.ferdjaoui@industrialdiscipline.com","username":"sahid"},"change_message_id":"63dce05eb4be0b9e8e2d4befea3cc9b05e98520c","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    elif disk_bus \u003d\u003d \"virtio\":"},{"line_number":162,"context_line":"        return 4194304"},{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":164,"context_line":"        return 256"},{"line_number":165,"context_line":"    else:"}],"source_content_type":"text/x-python","patch_set":18,"id":"5f7c97a3_81dafd1a","line":162,"range":{"start_line":162,"start_character":15,"end_line":162,"end_character":22},"updated":"2018-06-04 09:21:37.000000000","message":"blk_device_name() does not handle such value and we probably don\u0027t want a guest to be attached to such number. It seems that we don\u0027t have quota for that. So I would suggest to set a reasonable number.\n\nThat one will be a technical limit where then at some point we should be able to add for that quota too. What about 256 or 512.\n\nCan you add a TODO note explaining that we currently have hard-coded a limit and expect that a some point \u0027quota\u0027 would handle limiting number of volume attached per VM?","commit_id":"2ee20418f410ba257f005a2406e6e51a1a55f515"},{"author":{"_account_id":7730,"name":"Sahid Orentino Ferdjaoui","email":"sahid.ferdjaoui@industrialdiscipline.com","username":"sahid"},"change_message_id":"d123233d72d545010144af4f4abb3674f6909f5e","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    elif disk_bus \u003d\u003d \"virtio\":"},{"line_number":162,"context_line":"        return 4194304"},{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":164,"context_line":"        return 256"},{"line_number":165,"context_line":"    else:"}],"source_content_type":"text/x-python","patch_set":18,"id":"5f7c97a3_8860da0f","line":162,"range":{"start_line":162,"start_character":15,"end_line":162,"end_character":22},"in_reply_to":"5f7c97a3_81dafd1a","updated":"2018-06-04 13:41:21.000000000","message":"(An other point is that, in Nova for virtio-scsi we don\u0027t increment the target. So we have only one controller and we use one target. It\u0027s means we are more about 2^14 LUN available.)","commit_id":"2ee20418f410ba257f005a2406e6e51a1a55f515"},{"author":{"_account_id":7730,"name":"Sahid Orentino Ferdjaoui","email":"sahid.ferdjaoui@industrialdiscipline.com","username":"sahid"},"change_message_id":"63dce05eb4be0b9e8e2d4befea3cc9b05e98520c","unresolved":false,"context_lines":[{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    elif disk_bus \u003d\u003d \"virtio\":"},{"line_number":162,"context_line":"        return 4194304"},{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":164,"context_line":"        return 256"},{"line_number":165,"context_line":"    else:"},{"line_number":166,"context_line":"        return 26"}],"source_content_type":"text/x-python","patch_set":18,"id":"5f7c97a3_e14971b7","line":163,"updated":"2018-06-04 09:21:37.000000000","message":"I would have make that change related to the native QEMU scsi implementation in a separated patch.","commit_id":"2ee20418f410ba257f005a2406e6e51a1a55f515"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"c2c4086643bc569d3a834a15efa0f42c70b44927","unresolved":false,"context_lines":[{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    elif disk_bus \u003d\u003d \"virtio\":"},{"line_number":162,"context_line":"        return 4194304"},{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":164,"context_line":"        return 256"},{"line_number":165,"context_line":"    else:"},{"line_number":166,"context_line":"        return 26"}],"source_content_type":"text/x-python","patch_set":18,"id":"5f7c97a3_0219fd20","line":163,"in_reply_to":"5f7c97a3_e14971b7","updated":"2018-06-05 00:23:34.000000000","message":"thx! sahid","commit_id":"2ee20418f410ba257f005a2406e6e51a1a55f515"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"d2182ddf726a537ad93e4a3f357887b9a17d2ce1","unresolved":false,"context_lines":[{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":164,"context_line":"        return 256"},{"line_number":165,"context_line":"    else:"},{"line_number":166,"context_line":"        return 26"},{"line_number":167,"context_line":""},{"line_number":168,"context_line":""},{"line_number":169,"context_line":"def blk_device_name(prefix, index):"}],"source_content_type":"text/x-python","patch_set":18,"id":"5f7c97a3_4b83b2cb","line":166,"range":{"start_line":166,"start_character":15,"end_line":166,"end_character":17},"updated":"2018-06-02 00:29:37.000000000","message":"historically, 26 is a one alphabet limitation, its should be more convenient value, like 128.","commit_id":"2ee20418f410ba257f005a2406e6e51a1a55f515"},{"author":{"_account_id":7730,"name":"Sahid Orentino Ferdjaoui","email":"sahid.ferdjaoui@industrialdiscipline.com","username":"sahid"},"change_message_id":"63dce05eb4be0b9e8e2d4befea3cc9b05e98520c","unresolved":false,"context_lines":[{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":164,"context_line":"        return 256"},{"line_number":165,"context_line":"    else:"},{"line_number":166,"context_line":"        return 26"},{"line_number":167,"context_line":""},{"line_number":168,"context_line":""},{"line_number":169,"context_line":"def blk_device_name(prefix, index):"}],"source_content_type":"text/x-python","patch_set":18,"id":"5f7c97a3_c1f0f594","line":166,"range":{"start_line":166,"start_character":15,"end_line":166,"end_character":17},"in_reply_to":"5f7c97a3_4b83b2cb","updated":"2018-06-04 09:21:37.000000000","message":"I don\u0027t think so as indicated here:\n\nhttps://review.openstack.org/#/c/567472/16/nova/virt/libvirt/blockinfo.py@162\n\nMy thinking is that you could add a TODO note explaining that some checks need to be done based on the kernel/QEMU version so we would be able to increase the value.","commit_id":"2ee20418f410ba257f005a2406e6e51a1a55f515"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"c2c4086643bc569d3a834a15efa0f42c70b44927","unresolved":false,"context_lines":[{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":164,"context_line":"        return 256"},{"line_number":165,"context_line":"    else:"},{"line_number":166,"context_line":"        return 26"},{"line_number":167,"context_line":""},{"line_number":168,"context_line":""},{"line_number":169,"context_line":"def blk_device_name(prefix, index):"}],"source_content_type":"text/x-python","patch_set":18,"id":"5f7c97a3_221679f3","line":166,"range":{"start_line":166,"start_character":15,"end_line":166,"end_character":17},"in_reply_to":"5f7c97a3_c1f0f594","updated":"2018-06-05 00:23:34.000000000","message":"Done","commit_id":"2ee20418f410ba257f005a2406e6e51a1a55f515"},{"author":{"_account_id":15888,"name":"Zhenyu Zheng","email":"zheng.zhenyu@outlook.com","username":"Kevin_Zheng"},"change_message_id":"3315760f429a7f63cc9ae678211b194770270148","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    elif disk_bus \u003d\u003d \"virtio\":"},{"line_number":162,"context_line":"        return 4194304"},{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":164,"context_line":"        return 256"},{"line_number":165,"context_line":"    else:"}],"source_content_type":"text/x-python","patch_set":22,"id":"5f7c97a3_8e82b8a4","line":162,"range":{"start_line":162,"start_character":8,"end_line":162,"end_character":22},"updated":"2018-06-08 08:58:25.000000000","message":"this seems really unrealistic to me...","commit_id":"df863b53bf26699174d011e02d6bddf22e536760"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"f4de89145b26429a215cd830be7af8791678f53a","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    elif disk_bus \u003d\u003d \"virtio\":"},{"line_number":162,"context_line":"        return 4194304"},{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":164,"context_line":"        return 256"},{"line_number":165,"context_line":"    else:"}],"source_content_type":"text/x-python","patch_set":22,"id":"3f79a3b5_b9e75eac","line":162,"range":{"start_line":162,"start_character":8,"end_line":162,"end_character":22},"in_reply_to":"3f79a3b5_d51f9658","updated":"2018-08-27 07:27:09.000000000","message":"virtio(virtio-scsi) takes 2^14 devices. The source is sahid\u0027s comment.  https://review.openstack.org/#/c/567472/18/nova/virt/libvirt/blockinfo.py@162\nbut i\u0027llchange it.","commit_id":"df863b53bf26699174d011e02d6bddf22e536760"},{"author":{"_account_id":25562,"name":"Chen","email":"dstbtgagt@foxmail.com","username":"chenn2"},"change_message_id":"dc9f9249c9c49cd9a952c7afcc67b140b5de8848","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    elif disk_bus \u003d\u003d \"virtio\":"},{"line_number":162,"context_line":"        return 4194304"},{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":164,"context_line":"        return 256"},{"line_number":165,"context_line":"    else:"}],"source_content_type":"text/x-python","patch_set":22,"id":"3f79a3b5_d51f9658","line":162,"range":{"start_line":162,"start_character":8,"end_line":162,"end_character":22},"in_reply_to":"5f7c97a3_8e82b8a4","updated":"2018-08-24 15:28:11.000000000","message":"Agree unless a source indicating this limit is provided.","commit_id":"df863b53bf26699174d011e02d6bddf22e536760"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"f4de89145b26429a215cd830be7af8791678f53a","unresolved":false,"context_lines":[{"line_number":159,"context_line":"    if disk_bus \u003d\u003d \"ide\":"},{"line_number":160,"context_line":"        return 4"},{"line_number":161,"context_line":"    elif disk_bus \u003d\u003d \"virtio\":"},{"line_number":162,"context_line":"        return 4194304"},{"line_number":163,"context_line":"    elif disk_bus \u003d\u003d \"scsi\":"},{"line_number":164,"context_line":"        return 256"},{"line_number":165,"context_line":"    else:"}],"source_content_type":"text/x-python","patch_set":22,"id":"3f79a3b5_d99fbaff","line":162,"range":{"start_line":162,"start_character":8,"end_line":162,"end_character":22},"in_reply_to":"5f7c97a3_8e82b8a4","updated":"2018-08-27 07:27:09.000000000","message":"how about 1k. Can you agree?\nThis number comes from sahid\u0027s comment.\nhttps://review.openstack.org/#/c/567472/18/nova/virt/libvirt/blockinfo.py@162","commit_id":"df863b53bf26699174d011e02d6bddf22e536760"}],"releasenotes/notes/bp-nova-improvement-of-maximum-attach-volumes-more-than-26-vols-4e9f7db3ad50a606.yaml":[{"author":{"_account_id":4690,"name":"melanie witt","display_name":"melwitt","email":"melwittt@gmail.com","username":"melwitt"},"change_message_id":"5b63d28149abc80fd656638eb2d9441c48018dc6","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"features:"},{"line_number":3,"context_line":"  - Improvement of maximum attach volumes more than"},{"line_number":4,"context_line":"    26 volumes each instances."}],"source_content_type":"text/x-yaml","patch_set":11,"id":"5f7c97a3_780c273b","line":4,"updated":"2018-05-17 16:48:04.000000000","message":"Suggestion: Increase the maximum allowed number of volumes attached to a single instance from 26 to N for \u0027virtio\u0027 and \u0027scsi\u0027.\n\n(replacing N with the new lower proposed max we choose)","commit_id":"576f360ac9318e230fc31e58f88532c8edaf461d"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"de1b94d861d2cf6721695da90f37f1cfd146c2e9","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"features:"},{"line_number":3,"context_line":"  - Improvement of maximum attach volumes more than"},{"line_number":4,"context_line":"    26 volumes each instances."}],"source_content_type":"text/x-yaml","patch_set":11,"id":"5f7c97a3_c2f1bc03","line":4,"in_reply_to":"5f7c97a3_780c273b","updated":"2018-05-18 03:55:33.000000000","message":"In the case of small number, I changed original code \u00274\u0027 or \u0027N\u0027","commit_id":"576f360ac9318e230fc31e58f88532c8edaf461d"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"8cfa106619dd947b465819d7d79d200eaf8a536f","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"features:"},{"line_number":3,"context_line":"  - Increase the maximum allowed number of volumes attached to"},{"line_number":4,"context_line":"    a single instance from 26 to 64."},{"line_number":5,"context_line":""}],"source_content_type":"text/x-yaml","patch_set":13,"id":"5f7c97a3_61f4545c","line":4,"updated":"2018-05-18 22:59:24.000000000","message":"from 26 to 256 for \u0027virtio\u0027","commit_id":"10ebc2eedd644bc10bb4e7c6b70b5e93a3f6400c"},{"author":{"_account_id":15888,"name":"Zhenyu Zheng","email":"zheng.zhenyu@outlook.com","username":"Kevin_Zheng"},"change_message_id":"ab7c3876930053017af6f69214a41f5ed1dbc7c0","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"features:"},{"line_number":3,"context_line":"  - Increase the maximum allowed number of volumes attached to"},{"line_number":4,"context_line":"    a single instance from 26 to 128."},{"line_number":5,"context_line":""}],"source_content_type":"text/x-yaml","patch_set":15,"id":"5f7c97a3_52d79a50","line":4,"range":{"start_line":3,"start_character":4,"end_line":4,"end_character":37},"updated":"2018-05-21 06:55:48.000000000","message":"Should we mention disk bus here? as the limit for IDE is still 4","commit_id":"f43b85c4f96021113ed285b6306a0345c10bf784"},{"author":{"_account_id":15888,"name":"Zhenyu Zheng","email":"zheng.zhenyu@outlook.com","username":"Kevin_Zheng"},"change_message_id":"f30349bf9c0ae38f812442e562835cecb1d9ebe7","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"features:"},{"line_number":3,"context_line":"  - Increase the maximum allowed number of volumes attached to"},{"line_number":4,"context_line":"    a single instance from 26 to 128."},{"line_number":5,"context_line":""}],"source_content_type":"text/x-yaml","patch_set":15,"id":"5f7c97a3_466cbb35","line":4,"range":{"start_line":3,"start_character":4,"end_line":4,"end_character":37},"in_reply_to":"5f7c97a3_26997f78","updated":"2018-05-21 08:57:11.000000000","message":"Thanks for the clearification, I\u0027m not aware of that IDE are legacy things, I\u0027d say this is a soft -1, just nice to add if you have to upload a new patchset, lets see how others think about this. :)","commit_id":"f43b85c4f96021113ed285b6306a0345c10bf784"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"c1ffecad10090df0c37938433ea06dcaa738a9ab","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"features:"},{"line_number":3,"context_line":"  - Increase the maximum allowed number of volumes attached to"},{"line_number":4,"context_line":"    a single instance from 26 to 128."},{"line_number":5,"context_line":""}],"source_content_type":"text/x-yaml","patch_set":15,"id":"5f7c97a3_0a3eb428","line":4,"range":{"start_line":3,"start_character":4,"end_line":4,"end_character":37},"in_reply_to":"5f7c97a3_466cbb35","updated":"2018-05-22 12:49:46.000000000","message":"Done","commit_id":"f43b85c4f96021113ed285b6306a0345c10bf784"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"3bf749d8914481cec43a02a9d2b10a230ca4939d","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"features:"},{"line_number":3,"context_line":"  - Increase the maximum allowed number of volumes attached to"},{"line_number":4,"context_line":"    a single instance from 26 to 128."},{"line_number":5,"context_line":""}],"source_content_type":"text/x-yaml","patch_set":15,"id":"5f7c97a3_26997f78","line":4,"range":{"start_line":3,"start_character":4,"end_line":4,"end_character":37},"in_reply_to":"5f7c97a3_52d79a50","updated":"2018-05-21 08:38:17.000000000","message":"No. No one use attaching IDE volumes on latest openstack.\nIs it usefull for legacy user?\n\u003e a single instance from 26 to 128 (except 4-IDE limits)","commit_id":"f43b85c4f96021113ed285b6306a0345c10bf784"},{"author":{"_account_id":7730,"name":"Sahid Orentino Ferdjaoui","email":"sahid.ferdjaoui@industrialdiscipline.com","username":"sahid"},"change_message_id":"4c3779b624730586ebe2500efbcbaeb55f959549","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"features:"},{"line_number":3,"context_line":"  - Increase the maximum allowed number of volumes attached to"},{"line_number":4,"context_line":"    a single instance from 26 to 128 (except for IDE is stil 4)."}],"source_content_type":"text/x-yaml","patch_set":17,"id":"5f7c97a3_69a8048b","line":3,"updated":"2018-06-01 13:46:40.000000000","message":"You need to update that.","commit_id":"d2f151b311a8e95a8833e37f517a0ae83225d7a1"},{"author":{"_account_id":28203,"name":"Tsuyoshi Nagata","email":"nagata3333333@jp.fujitsu.com","username":"yukari-papa"},"change_message_id":"ec335164d18bd541abdaa520807008e53e0faac1","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"features:"},{"line_number":3,"context_line":"  - Increase the maximum allowed number of volumes attached to"},{"line_number":4,"context_line":"    a single instance from 26 to 128 (except for IDE is stil 4)."}],"source_content_type":"text/x-yaml","patch_set":17,"id":"5f7c97a3_098b90bf","line":3,"in_reply_to":"5f7c97a3_69a8048b","updated":"2018-06-01 15:00:59.000000000","message":"Done","commit_id":"d2f151b311a8e95a8833e37f517a0ae83225d7a1"}]}