)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":14119,"name":"Sam Yaple","email":"sam+git@yaple.net","username":"SamYaple"},"change_message_id":"d1e77ce9daabe686e0434ae9f00cfb2f88d91b7e","unresolved":false,"context_lines":[{"line_number":8,"context_line":""},{"line_number":9,"context_line":"The existing gate partitioned a disk for use with docker, depending"},{"line_number":10,"context_line":"on the gate it would use the swap disk (RAX) or a spare disk (HP)."},{"line_number":11,"context_line":"However, with the new gates (Bluebox + OVH) there is neither a spare"},{"line_number":12,"context_line":"disk nor a swap disk. This leaves us with one choice: File based loop"},{"line_number":13,"context_line":"device."},{"line_number":14,"context_line":""}],"source_content_type":"text/x-gerrit-commit-message","patch_set":24,"id":"9a68dd71_7ff1386a","line":11,"updated":"2016-01-25 15:08:45.000000000","message":"Some gates are instances hosted in RAX, some HP, some Bluebox, and some OVH. These all have little differences in vm harware and flavor. There is no migration for a running gate, but a new patch or a recheck means it could be on a different provider","commit_id":"6adb5c0aa5a5947bdeeb03b03cba29af30902d2f"},{"author":{"_account_id":14119,"name":"Sam Yaple","email":"sam+git@yaple.net","username":"SamYaple"},"change_message_id":"14802e8b0df92e84186eb4aed258eae36a0e7c3a","unresolved":false,"context_lines":[{"line_number":8,"context_line":""},{"line_number":9,"context_line":"The existing gate partitioned a disk for use with docker, depending"},{"line_number":10,"context_line":"on the gate it would use the swap disk (RAX) or a spare disk (HP)."},{"line_number":11,"context_line":"However, with the new gates (Bluebox + OVH) there is neither a spare"},{"line_number":12,"context_line":"disk nor a swap disk. This leaves us with one choice: File based loop"},{"line_number":13,"context_line":"device."},{"line_number":14,"context_line":""}],"source_content_type":"text/x-gerrit-commit-message","patch_set":24,"id":"9a68dd71_c76655ec","line":11,"updated":"2016-01-25 17:26:46.000000000","message":"anytime Jeffrey :)","commit_id":"6adb5c0aa5a5947bdeeb03b03cba29af30902d2f"},{"author":{"_account_id":7488,"name":"Jeffrey Zhang","email":"zhang.lei.fly@gmail.com","username":"Jeffrey4l"},"change_message_id":"2160bd2e0d671044b63b2f1d4f98266710fd41c7","unresolved":false,"context_lines":[{"line_number":8,"context_line":""},{"line_number":9,"context_line":"The existing gate partitioned a disk for use with docker, depending"},{"line_number":10,"context_line":"on the gate it would use the swap disk (RAX) or a spare disk (HP)."},{"line_number":11,"context_line":"However, with the new gates (Bluebox + OVH) there is neither a spare"},{"line_number":12,"context_line":"disk nor a swap disk. This leaves us with one choice: File based loop"},{"line_number":13,"context_line":"device."},{"line_number":14,"context_line":""}],"source_content_type":"text/x-gerrit-commit-message","patch_set":24,"id":"9a68dd71_2845e744","line":11,"updated":"2016-01-25 01:09:12.000000000","message":"hey Sam, could u explain more about the RAX, HP, Bluebox + OVH? I have no idea about the jenkins env.\n\nDoes the job runs on both of there environment(random choice? configurable?), or migrate from the first two too the last one?","commit_id":"6adb5c0aa5a5947bdeeb03b03cba29af30902d2f"},{"author":{"_account_id":14119,"name":"Sam Yaple","email":"sam+git@yaple.net","username":"SamYaple"},"change_message_id":"b48ebf57ec23438b5893318daee4df93deece279","unresolved":false,"context_lines":[{"line_number":8,"context_line":""},{"line_number":9,"context_line":"The existing gate partitioned a disk for use with docker, depending"},{"line_number":10,"context_line":"on the gate it would use the swap disk (RAX) or a spare disk (HP)."},{"line_number":11,"context_line":"However, with the new gates (Bluebox + OVH) there is neither a spare"},{"line_number":12,"context_line":"disk nor a swap disk. This leaves us with one choice: File based loop"},{"line_number":13,"context_line":"device."},{"line_number":14,"context_line":""}],"source_content_type":"text/x-gerrit-commit-message","patch_set":24,"id":"9a68dd71_ca165f4b","line":11,"in_reply_to":"9a68dd71_24597435","updated":"2016-01-25 16:23:59.000000000","message":"Then you aren\u0027t looking at our gates very closely. I would say 10-20% of our gates fail for reasons that are addressed in this patchset.","commit_id":"6adb5c0aa5a5947bdeeb03b03cba29af30902d2f"},{"author":{"_account_id":7488,"name":"Jeffrey Zhang","email":"zhang.lei.fly@gmail.com","username":"Jeffrey4l"},"change_message_id":"a182e32c2a826684b22a6433be2c922c2c025a0f","unresolved":false,"context_lines":[{"line_number":8,"context_line":""},{"line_number":9,"context_line":"The existing gate partitioned a disk for use with docker, depending"},{"line_number":10,"context_line":"on the gate it would use the swap disk (RAX) or a spare disk (HP)."},{"line_number":11,"context_line":"However, with the new gates (Bluebox + OVH) there is neither a spare"},{"line_number":12,"context_line":"disk nor a swap disk. This leaves us with one choice: File based loop"},{"line_number":13,"context_line":"device."},{"line_number":14,"context_line":""}],"source_content_type":"text/x-gerrit-commit-message","patch_set":24,"id":"9a68dd71_24597435","line":11,"in_reply_to":"9a68dd71_7ff1386a","updated":"2016-01-25 15:54:12.000000000","message":"But I don\u0027t see errors without this PS. What\u0027s the deep reason for this? better performance or?","commit_id":"6adb5c0aa5a5947bdeeb03b03cba29af30902d2f"},{"author":{"_account_id":7488,"name":"Jeffrey Zhang","email":"zhang.lei.fly@gmail.com","username":"Jeffrey4l"},"change_message_id":"b03de0418cf931d385ad9a785f70bd5aa39836d7","unresolved":false,"context_lines":[{"line_number":8,"context_line":""},{"line_number":9,"context_line":"The existing gate partitioned a disk for use with docker, depending"},{"line_number":10,"context_line":"on the gate it would use the swap disk (RAX) or a spare disk (HP)."},{"line_number":11,"context_line":"However, with the new gates (Bluebox + OVH) there is neither a spare"},{"line_number":12,"context_line":"disk nor a swap disk. This leaves us with one choice: File based loop"},{"line_number":13,"context_line":"device."},{"line_number":14,"context_line":""}],"source_content_type":"text/x-gerrit-commit-message","patch_set":24,"id":"9a68dd71_356f5cfd","line":11,"in_reply_to":"9a68dd71_ca165f4b","updated":"2016-01-25 16:51:29.000000000","message":"OK. Understood. thanks a lot.","commit_id":"6adb5c0aa5a5947bdeeb03b03cba29af30902d2f"}],"tools/setup_RedHat.sh":[{"author":{"_account_id":7488,"name":"Jeffrey Zhang","email":"zhang.lei.fly@gmail.com","username":"Jeffrey4l"},"change_message_id":"1a04ec730816f4e9f06244d4806d9c76dda9071e","unresolved":false,"context_lines":[{"line_number":10,"context_line":"    sudo mkswap /swapfile"},{"line_number":11,"context_line":"    sudo /sbin/swapon /swapfile"},{"line_number":12,"context_line":""},{"line_number":13,"context_line":"    sudo dd if\u003d/dev/zero of\u003d/docker bs\u003d1M count\u003d10240"},{"line_number":14,"context_line":"    losetup -f /docker"},{"line_number":15,"context_line":"    DEV\u003d$(losetup -a | awk -F: \u0027/\\/docker/ {print $1}\u0027)"},{"line_number":16,"context_line":""}],"source_content_type":"text/x-sh","patch_set":24,"id":"9a68dd71_8fdb827b","line":13,"updated":"2016-01-26 01:43:29.000000000","message":"Do you ever consider using qemu-img and qemu-nbd? it doesn\u0027t require to write the whole disk. (maybe better performance? Just guess, no test data)","commit_id":"6adb5c0aa5a5947bdeeb03b03cba29af30902d2f"},{"author":{"_account_id":14119,"name":"Sam Yaple","email":"sam+git@yaple.net","username":"SamYaple"},"change_message_id":"aa46d711dbd69eaef785717f3af63810b48ec8b9","unresolved":false,"context_lines":[{"line_number":10,"context_line":"    sudo mkswap /swapfile"},{"line_number":11,"context_line":"    sudo /sbin/swapon /swapfile"},{"line_number":12,"context_line":""},{"line_number":13,"context_line":"    sudo dd if\u003d/dev/zero of\u003d/docker bs\u003d1M count\u003d10240"},{"line_number":14,"context_line":"    losetup -f /docker"},{"line_number":15,"context_line":"    DEV\u003d$(losetup -a | awk -F: \u0027/\\/docker/ {print $1}\u0027)"},{"line_number":16,"context_line":""}],"source_content_type":"text/x-sh","patch_set":24,"id":"9a68dd71_2a9284f6","line":13,"updated":"2016-01-26 02:50:57.000000000","message":"Using qemu-img/nbd will never result in _better_ performance than a raw fully allocated file than this. We might gain a bit of speed from simply not allocating all of this out, but its a 10 second operation so im not too worried. If we had more ram I would do bcache to increase performance. In fact I plan on a patchset to test just that!","commit_id":"6adb5c0aa5a5947bdeeb03b03cba29af30902d2f"}]}