minibug-MiniBug Sample Project.json

{"Version":"1.0","IssueIdCounter":11,"TaskIdCounter":6,"Name":"MiniBug Sample Project","Issues":{"1":{"ID":1,"Priority":2,"Status":1,"Version":"48 Branch","TargetVersion":"","Summary":"CSS layouterror: nth-child(even) is not calculated correctly","Description":"User Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:48.0) Gecko/20100101 Firefox/48.0\r\nBuild ID: 20160819105012\r\n\r\nSteps to reproduce:\r\n\r\nI have this piece of code :(ignore the missing ids)\r\n\r\n\r\n\r\n<div class=leftnavi>\r\n\t<div class=navbutton id=userlist>Benutzerliste</div>\r\n\t<div class=navbutton id=></div>\r\n\t<div class=navbutton id=></div>\r\n\t<div class=navbutton id=></div>\r\n\t<div class=navbutton id=></div>\r\n\t<br>\r\n\t<div class=navbutton id=>Webseite</div>\r\n\t<br>\r\n\t<div class=navbutton id=>Ausloggen</div>\r\n</div>\r\n\r\n\r\nwith this CSS:\r\n\r\n.leftnavi {\r\n  background: white none repeat scroll 0 0;\r\n  border-bottom: 1px solid black;\r\n  border-right: 1px solid black;\r\n  display: inline;\r\n  float: left;\r\n  padding-top: 1em;\r\n  padding-bottom: 1em;\r\n  border-radius: 0 0 1em 0;\r\n  padding-bottom: 1em;\r\n}\r\n\r\nand here comes the error : \r\n\r\n.leftnavi > DIV:nth-child(odd) {\r\n\tbackground: #eaeaea;\r\n}\r\n\r\n\r\nActual results:\r\n\r\nthe \"even\" and \"odd\" DIV's are miscalculated.\r\n\r\nThe css rule clearly referes to an even or odd DIV, but instead all elements are counted, this includes the <br> elements. This leads to the situation, that the last two DIV get the same background color, which is not what was intended.\r\n\r\n\r\n\r\n\r\nExpected results:\r\n\r\nthe <br> are not refered to, so they do not count for the even or odd calculation of those DIV's. The last and the second last DIV should end up with different background colors.","DateCreated":"2019-05-08T18:16:04.2898432+01:00","DateModified":"2019-05-08T18:16:04.2918285+01:00"},"2":{"ID":2,"Priority":5,"Status":1,"Version":"58 Branch","TargetVersion":"","Summary":"Browser hangs on infinite downloads of blobs","Description":"User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/64.0.3282.167 Safari/537.36\r\n\r\nSteps to reproduce:\r\n\r\nTested on Firefox 58.0.2 (64-bit) on Windows 7 64bit\r\n\r\nLoad the attached PoC.\r\n\r\n\r\nActual results:\r\n\r\nFirefox hangs with CPU at 100% and eventually memory runs out.\r\n\r\n\r\nExpected results:\r\n\r\nFirefox should prevent this behaviour and also allow the user to exit the tab or browser.\r\nThis technique is already used in the wild (but originally targeted Chrome users): https://blog.malwarebytes.com/malwarebytes-news/2018/02/tech-support-scammers-find-new-way-jam-google-chrome/","DateCreated":"2019-05-08T18:21:12.3250722+01:00","DateModified":"2019-05-08T18:21:12.3250722+01:00"},"3":{"ID":3,"Priority":5,"Status":2,"Version":"10.1.36","TargetVersion":"10.1","Summary":"Replication lag issue using parallel replication","Description":"Using parallel replication second behind master is wrongly reporting 0 when SQL thread is stopped and restarted long time after.\r\n\r\nThis happen by design\r\nhttps://lists.launchpad.net/maria-developers/msg08958.html\r\n\r\nbut is really a show stopper for most proxy that send traffic to such slave thinking it's in sync with master.\r\n\r\nMy understanding is that slave_behind_master is computed after first commit so in this case the master is 2 days in advance and on a fresh restarted slave we get this\r\n\r\n|   30 | system user  |                      | tsce_unedic | Connect | 2211 | altering table                                                                 | OPTIMIZE TABLE `requetes` |    0.000 |\r\n\r\nAnd we can see wrong second behind master\r\n\r\n        Seconds_Behind_Master: 0\r\n\r\n                   Using_Gtid: Slave_Pos\r\n\r\n                  Gtid_IO_Pos: 0-21-28557589\r\n\r\n                Parallel_Mode: conservative \r\n\r\nbut on his master\r\n\r\ngtid_current_pos       | 0-21-28570301 \r\n\r\nA possible solution would be to update Seconds_Behind_Master by injecting a fake event in start slave with the max timestamp of all events read by the leader thread and send to to the worker threads .\r\n\r\nTo reproduce :\r\n\r\n--source include/have_innodb.inc\r\n\r\n--source include/have_binlog_format_mixed.inc\r\n\r\n--let $rpl_topology=1->2\r\n\r\n--source include/rpl_init.inc\r\n\r\n \r\n\r\n# Test various aspects of parallel replication.\r\n\r\n \r\n\r\n--connection server_1\r\n\r\nALTER TABLE mysql.gtid_slave_pos ENGINE=InnoDB;\r\n\r\n \r\n\r\nCREATE TABLE t1 (a INT PRIMARY KEY, b INT) ENGINE=InnoDB;\r\n\r\n--save_master_pos\r\n\r\n \r\n\r\n--connection server_2\r\n\r\n--sync_with_master\r\n\r\n--source include/stop_slave_sql_thread.inc\r\n\r\nSET GLOBAL slave_parallel_threads=4;\r\n\r\n \r\n\r\n--connection server_2\r\n\r\n--sync_with_master\r\n\r\n--source include/stop_slave.inc\r\n\r\nSET GLOBAL slave_parallel_threads=1;\r\n\r\n \r\n\r\n--connection server_1\r\n\r\n--disable_warnings\r\n\r\nINSERT INTO t1 VALUES (1, SLEEP(100));\r\n\r\n--wait 100s\r\n\r\nINSERT INTO t1 VALUES (1, SLEEP(1));\r\n\r\n \r\n\r\n--connection server_2\r\n\r\n--source include/start_slave.inc\r\n\r\n--let $status_items= Seconds_Behind_Master\r\n\r\n--source include/show_slave_status.inc\r\n\r\n--sync_with_master\r\n\r\n--let $status_items= Seconds_Behind_Master\r\n\r\n--source include/show_slave_status.inc\r\n","DateCreated":"2019-05-08T20:44:13.114665+01:00","DateModified":"2019-05-08T20:44:13.114665+01:00"},"4":{"ID":4,"Priority":5,"Status":4,"Version":"","TargetVersion":"Inkscape 1.0","Summary":" Mousewheel scroll and zoom broken on Win7","Description":"I cannot use the mousewheel or Ctrl-mousewheel in the document in the latest inkscape appveyor Windows build on Win7 64bit (inside VirtualBox, both with the default virtual USB tablet mouse or with a virtual PS2 mouse).\r\n\r\nI bisected it to:\r\nOn Windows 7 64bit\r\n- good up to (including) inkscape-0.92+devel_2018-08-21_d890608-x64.7z\r\n- unknown inbetween (no appveyor builds, will compile manually and see where exactly it broke)\r\n- broken since (including) inkscape-0.92+devel_2018-09-07_b9177a8-x64.7z\r\n\r\nOn Linux everything works.\r\n\r\nIt may also be due to a change in the appveyor build system.","DateCreated":"2019-05-08T21:32:07.6041441+01:00","DateModified":"2019-05-08T21:32:07.6041441+01:00"},"5":{"ID":5,"Priority":1,"Status":2,"Version":"v18 Leia","TargetVersion":"","Summary":"JSON API Player.Seek: Negative Time Interval Causes Incorrect Behavior","Description":"Here is a clear and concise description of what the problem is:\r\n\r\nUse of negative \"seconds\" time unit in Player.Seek JSON command results in unexpected behavior that is inconsistent with every other behavior pattern from Player.Seek\r\nExpected Behavior\r\n\r\nHere is a clear and concise description of what was expected to happen:\r\n\r\nError message. Any time interval other than seconds produces an error message, as expected. Player.Seek should not hand off the properties to another process.\r\n\r\nIf the user enters a negative minutes, hours, or milliseconds property, an error message is returned: \"Received value does not match any of the union type definitions\"\r\nActual Behavior\r\n\r\nProduces erroneous behavior inconsistent with Player.Seek expected behavior. The media player acts as if it received a Rewind command.\r\n\r\nExample:\r\n\r\n{\"jsonrpc\": \"2.0\", \"method\": \"Player.Seek\", \"params\":{\"playerid\":1,\"value\":{\"seconds\":-15}},\"id\": 1}\r\n\r\nThe result in this use case is similar to Rewind (re: Input.Action). Player.Seek appears to use the incorrect time index, to which the negative \"seconds\" value is applied. The time offset appears to be indexed (incorrectly) from the current time instead of time 0 (zero; i.e. beginning of media). The latter pattern is consistent with all other behavior from Player.Seek\r\nPossible Fix\r\n\r\nThis only occurs under one scenario. When all of the following are true:\r\n\r\n    Player ID is valid\r\n    Only the Seconds time unit is received\r\n    Unit time Value is a negative integer\r\n\r\nFor example:\r\n\r\n{\"jsonrpc\": \"2.0\", \"method\": \"Player.Seek\", \"params\":{\"playerid\":1,\"value\":{\"minutes\":15, \"seconds\": 12}},\"id\": 1}\r\n\r\nI suspect one of two scenarios is occurring. Either\r\n\r\na) Code is branching incorrectly when Player.Seek receives the specified object properties\r\n\r\nOR\r\n\r\nb) The logic surrounding the time index, which is normally zero (0) or the beginning of the media content (e.g. movie) is not being zero'd out, but rather the current media player time pointer is being used as an index, to which the time value in the JSON command is then applied\r\nTo Reproduce\r\n\r\nSteps to reproduce the behavior:\r\n\r\n    Begin playing a video file in Kodi\r\n    Wait until the timer for the video is past the first minute\r\n    Using CURL, Postman, or similar test tool, send a JSON-RPC API command to Kodi, similar to the following example. The command should include a Player ID = 1 and the Seconds parameter equal to a negative integer less than 60. Suggest using a value greater than 10 to ensure the behavior is obvious.\r\n\r\n{\"jsonrpc\": \"2.0\", \"method\": \"Player.Seek\", \"params\":{\"playerid\":1,\"value\":{\"seconds\":-15}},\"id\": 1}\r\n\r\n    Observe that Kodi behaves as if it received a Rewind command (e.g. via Input.Action), and rewinds the timer pointing to the current time being played on the video media by the same number of seconds indicated in the JSON command.\r\n\r\nDebuglog\r\n\r\nThe debuglog can be found here:\r\nScreenshots\r\n\r\nHere are some links or screenshots to help explain the problem:\r\nAdditional context or screenshots (if appropriate)\r\n\r\nHere is some additional context or explanation that might help:\r\n\r\nHow do I know this is a bug and not WAD? Because if I use any other time unit (i.e. milliseconds, minutes, hours), I receive an error message. Furthermore, the opposite behavior (using a positive integer) results in a different set of behavior. Namely, Player.Seek moves the media player time pointer to an offset from the beginning of the media (start) equal to the time units passed in the JSON command.\r\n\r\nFor example, if I am watching a movie that is at time point 01:02:03 and I execute this command in the API:\r\n\r\n{\"jsonrpc\": \"2.0\", \"method\": \"Player.Seek\", \"params\":{\"playerid\":1,\"value\":{\"minutes\":15, \"seconds\": 12}},\"id\": 1}\r\n\r\nThe result is the media player moves the current pointer (point in the movie being played right now) to 15 minutes, 12 seconds from the beginning of of the movie.\r\n\r\nAnother example supporting my argument: If the same type of command as the error-producing command is implemented, except the Value integer is positive, Player.Seek treats the Value as a percentage and passes the instruction on to Player.Position.Percentage\r\n\r\n{\"jsonrpc\": \"2.0\", \"method\": \"Player.Seek\", \"params\":{\"playerid\":1,\"value\":17},\"id\": 1}\r\n\r\nThe result is the currently playing media pointer shifts the time to 17% of the total movie length. On a 2 hour long movie, that would be 20.4 minutes into the film.","DateCreated":"2019-05-08T21:56:46.3428505+01:00","DateModified":"2019-05-08T21:56:46.3428505+01:00"},"6":{"ID":6,"Priority":3,"Status":2,"Version":"0.48.4","TargetVersion":"","Summary":"Fill/Stroke crash","Description":"If I select a tool like the 'freehand lines' or 'calligraphy' and click on the fill/stroke icon in the corner of the screen, inkscape crashes with the message 'inkscape encountered an internal error and will close now'. I am using version 0.48.4 on windows 7 x64 and experience this consistently, whether I use a mouse or tablet.","DateCreated":"2019-05-08T22:03:11.1696139+01:00","DateModified":"2019-05-08T22:03:11.1696139+01:00"},"7":{"ID":7,"Priority":5,"Status":3,"Version":"10.1","TargetVersion":"10.2","Summary":"InnoDB: Failing assertion: ut_strcmp(index->name, key->name)","Description":"With our current setup (and also on other hardware, but same software versions)\r\nwe keep getting this issue. This due a process of running a process to create a index table.\r\n\r\n[Note] InnoDB: Index for key no 1 mysql name \r\n\r\n \r\n\r\n{code:java}\r\n\r\nCATALOG_PRODUCT_FLAT_1_TYPE_ID InnoDB name CATALOG_PRODUCT_FLAT_8_ATTRIBUTE_SET_ID for table app2/#sql-26be_424c6\r\n\r\n \r\n\r\nFor table application1/#sql-26be_424c6 \r\n\r\napr 25 13:07:10 db1 mysqld[9918]: 2018-04-25 13:07:10 0x7f8f35189700 InnoDB: Assertion failure in file /home/buildbot/buildbot/build/mariadb-10.2.14/storage/innobase/handler/ha_innodb.cc line 9631 \r\n\r\napr 25 13:07:10 db1 mysqld[9918]: InnoDB: Failing assertion: ut_strcmp(index->name, key->name) == 0 \r\n\r\napr 25 13:07:10 db1 mysqld[9918]: InnoDB: We intentionally generate a memory trap. \r\n\r\napr 25 13:07:10 db1 mysqld[9918]: InnoDB: Submit a detailed bug report to https://jira.mariadb.org/ \r\n\r\napr 25 13:07:10 db1 mysqld[9918]: InnoDB: If you get repeated assertion failures or crashes, even \r\n\r\napr 25 13:07:10 db1 mysqld[9918]: InnoDB: immediately after the mysqld startup, there may be \r\n\r\napr 25 13:07:10 db1 mysqld[9918]: InnoDB: corruption in the InnoDB tablespace. Please refer to \r\n\r\napr 25 13:07:10 db1 mysqld[9918]: InnoDB: https://mariadb.com/kb/en/library/xtradbinnodb-recovery-modes/ \r\n\r\napr 25 13:07:10 db1 mysqld[9918]: InnoDB: about forcing recovery. \r\n\r\napr 25 13:07:10 db1 mysqld[9918]: 180425 13:07:10 [ERROR] mysqld got signal 6 ; \r\n\r\napr 25 13:07:10 db1 mysqld[9918]: This could be because you hit a bug. It is also possible that this binary \r\n\r\napr 25 13:07:10 db1 mysqld[9918]: or one of the libraries it was linked against is corrupt, improperly built, \r\n\r\napr 25 13:07:10 db1 mysqld[9918]: or misconfigured. This error can also be caused by malfunctioning hardware. \r\n\r\napr 25 13:07:10 db1 mysqld[9918]: To report this bug, see https://mariadb.com/kb/en/reporting-bugs \r\n\r\napr 25 13:07:10 db1 mysqld[9918]: We will try our best to scrape up some info that will hopefully help \r\n\r\napr 25 13:07:10 db1 mysqld[9918]: diagnose the problem, but since we have already crashed, \r\n\r\napr 25 13:07:10 db1 mysqld[9918]:\r\n\r\nMariadb server dies after this, and we need to recover manually to get things running again.\r\n\r\n \r\n\r\n{code:java}\r\n\r\nMaridb my.cnf settings: \r\n\r\n{noformat}\r\n\r\n# MariaDB database server configuration file.\r\n\r\n#\r\n\r\n# You can copy this file to one of:\r\n\r\n# - \"/etc/mysql/my.cnf\" to set global options,\r\n\r\n# - \"~/.my.cnf\" to set user-specific options.\r\n\r\n# \r\n\r\n# One can use all long options that the program supports.\r\n\r\n# Run program with --help to get a list of available options and with\r\n\r\n# --print-defaults to see which it would actually understand and use.\r\n\r\n#\r\n\r\n# For explanations see\r\n\r\n# http://dev.mysql.com/doc/mysql/en/server-system-variables.html\r\n\r\n \r\n\r\n# This will be passed to all mysql clients\r\n\r\n# It has been reported that passwords should be enclosed with ticks/quotes\r\n\r\n# escpecially if they contain \"#\" chars...\r\n\r\n# Remember to edit /etc/mysql/debian.cnf when changing the socket location.\r\n\r\n[client]\r\n\r\nport            = 3306\r\n\r\nsocket          = /var/run/mysqld/mysqld.sock\r\n\r\n \r\n\r\n# Here is entries for some specific programs\r\n\r\n# The following values assume you have at least 32M ram\r\n\r\n \r\n\r\n# This was formally known as [safe_mysqld]. Both versions are currently parsed.\r\n\r\n[mysqld_safe]\r\n\r\nsocket          = /var/run/mysqld/mysqld.sock\r\n\r\nnice            = 0\r\n\r\n \r\n\r\n[mysqld]","DateCreated":"2019-05-08T22:09:10.4118617+01:00","DateModified":"2019-05-08T22:12:07.1513434+01:00"},"8":{"ID":8,"Priority":5,"Status":3,"Version":"10.2.22, 10.2.23 ","TargetVersion":"10.2","Summary":"DB crashes periodicly with - InnoDB: Assertion failure","Description":"The DB service stops from time to time but non-periodic (estimated mostly between 11pm and 9am) with the following error:\r\n\r\n2019-03-28 23:31:08 0x7ff4dc62d700  InnoDB: Assertion failure in file /home/buildbot/buildbot/build/mariadb-10.2.23/storage/innobase/rem/rem0rec.cc line 574\r\n\r\nInnoDB: We intentionally generate a memory trap.\r\n\r\nInnoDB: Submit a detailed bug report to https://jira.mariadb.org/\r\n\r\nInnoDB: If you get repeated assertion failures or crashes, even\r\n\r\nInnoDB: immediately after the mysqld startup, there may be\r\n\r\nInnoDB: corruption in the InnoDB tablespace. Please refer to\r\n\r\nInnoDB: https://mariadb.com/kb/en/library/innodb-recovery-modes/\r\n\r\nInnoDB: about forcing recovery.\r\n\r\n190328 23:31:08 [ERROR] mysqld got signal 6 ;\r\n\r\nThis could be because you hit a bug. It is also possible that this binary\r\n\r\nor one of the libraries it was linked against is corrupt, improperly built,\r\n\r\nor misconfigured. This error can also be caused by malfunctioning hardware.\r\n\r\n \r\n\r\nTo report this bug, see https://mariadb.com/kb/en/reporting-bugs\r\n\r\n \r\n\r\nWe will try our best to scrape up some info that will hopefully help\r\n\r\ndiagnose the problem, but since we have already crashed, \r\n\r\nsomething is definitely wrong and this may fail.\r\n\r\n \r\n\r\nServer version: 10.2.23-MariaDB-10.2.23+maria~stretch\r\n\r\nkey_buffer_size=134217728\r\n\r\nread_buffer_size=131072\r\n\r\nmax_used_connections=234\r\n\r\nmax_threads=752\r\n\r\nthread_count=273\r\n\r\nIt is possible that mysqld could use up to \r\n\r\nkey_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 1783435 K  bytes of memory\r\n\r\nHope that's ok; if not, decrease some variables in the equation.\r\n\r\n \r\n\r\nThread pointer: 0x7ff4d8001f28\r\n\r\nAttempting backtrace. You can use the following information to find out\r\n\r\nwhere mysqld died. If you see no messages after this, something went\r\n\r\nterribly wrong...\r\n\r\nstack_bottom = 0x7ff4dc62ccc8 thread_stack 0x49000\r\n\r\n*** buffer overflow detected ***: /usr/sbin/mysqld terminated\r\n\r\n======= Backtrace: =========\r\n\r\n/lib/x86_64-linux-gnu/libc.so.6(+0x70bfb)[0x7ffa09af5bfb]\r\n\r\n/lib/x86_64-linux-gnu/libc.so.6(__fortify_fail+0x37)[0x7ffa09b7e437]\r\n\r\n/lib/x86_64-linux-gnu/libc.so.6(+0xf7570)[0x7ffa09b7c570]\r\n\r\n/lib/x86_64-linux-gnu/libc.so.6(+0xf93aa)[0x7ffa09b7e3aa]\r\n\r\n/usr/sbin/mysqld(my_addr_resolve+0xe2)[0x55ca42284922]\r\n\r\n/usr/sbin/mysqld(my_print_stacktrace+0x1bb)[0x55ca4226b1eb]\r\n\r\n/usr/sbin/mysqld(handle_fatal_signal+0x41d)[0x55ca41d0a01d]\r\n\r\n/lib/x86_64-linux-gnu/libpthread.so.0(+0x110e0)[0x7ffa0b4180e0]\r\n\r\n/lib/x86_64-linux-gnu/libc.so.6(gsignal+0xcf)[0x7ffa09ab7fff]\r\n\r\n/lib/x86_64-linux-gnu/libc.so.6(abort+0x16a)[0x7ffa09ab942a]\r\n\r\n/usr/sbin/mysqld(+0x40f971)[0x55ca41ab8971]\r\n\r\n/usr/sbin/mysqld(+0x887df6)[0x55ca41f30df6]\r\n\r\n/usr/sbin/mysqld(+0x863673)[0x55ca41f0c673]\r\n\r\n/usr/sbin/mysqld(+0x96648e)[0x55ca4200f48e]\r\n\r\n/usr/sbin/mysqld(+0x89b559)[0x55ca41f44559]","DateCreated":"2019-05-08T22:16:00.5184435+01:00","DateModified":"2019-05-08T22:16:00.5184435+01:00"},"9":{"ID":9,"Priority":3,"Status":4,"Version":"10.1.24","TargetVersion":"10.4","Summary":"Query optimizer not picking optimal ORDER BY PRIMARY in case of INNER JOIN on PRIMARY like it does for similar WHERE condition","Description":"Order by a column that isn't the first column in the join list requires a temporary table.\r\n\r\nA query that INNER JOINs two tables on their Primary keys with a WHERE clause on a particular key value properly optimizes to target the first table in the join order for the WHERE clause, presumably based on the logical equivalence of the values implied by the INNER JOIN:\r\n\r\nQuery second table's id (PRIMARY):\r\n\r\nSELECT first.id FROM first INNER JOIN second ON first.id = second.id WHERE second.id > '5E1215B77BB14DA4DD7C9D4DDD26501A';\r\n\r\nOptimized query (from EXPLAIN EXTENDED warning) points the WHERE clause to the first table:\r\n\r\nselect `db`.`first`.`id` AS `id` from `db`.`first` join `db`.`second` where ((`db`.`second`.`id` = `db`.`first`.`id`) and (`db`.`first`.`id` > '5E1215B77BB14DA4DD7C9D4DDD26501A'))\r\n\r\nThe same thing does not happen for the ORDER BY, and so changing which table's Primary key we order by makes the difference between requiring a temporary table and not:\r\n\r\nThis query (ORDER BY first) is basically just an index retrieval:\r\n\r\n> ANALYZE SELECT first.id FROM first INNER JOIN second ON first.id = second.id ORDER BY first.id;\r\n\r\n+------+-------------+--------+--------+---------------+---------+---------+------------------------+------+--------+----------+------------+-------------+\r\n\r\n| id   | select_type | table  | type   | possible_keys | key     | key_len | ref                    | rows | r_rows | filtered | r_filtered | Extra       |\r\n\r\n+------+-------------+--------+--------+---------------+---------+---------+------------------------+------+--------+----------+------------+-------------+\r\n\r\n|    1 | SIMPLE      | first  | index  | PRIMARY       | PRIMARY | 96      | NULL                   |  100 | 100.00 |   100.00 |     100.00 | Using index |\r\n\r\n|    1 | SIMPLE      | second | eq_ref | PRIMARY       | PRIMARY | 96      | huff_20170614.first.id |    1 |   1.00 |   100.00 |     100.00 | Using index |\r\n\r\n+------+-------------+--------+--------+---------------+---------+---------+------------------------+------+--------+----------+------------+-------------+\r\n\r\nBut if you change the ORDER BY to the second table, it now requires a temporary table:\r\n\r\n> ANALYZE SELECT first.id FROM first INNER JOIN second ON first.id = second.id ORDER BY second.id;\r\n\r\n+------+-------------+--------+--------+---------------+---------+---------+------------------------+------+--------+----------+------------+----------------------------------------------+\r\n\r\n| id   | select_type | table  | type   | possible_keys | key     | key_len | ref                    | rows | r_rows | filtered | r_filtered | Extra                                        |\r\n\r\n+------+-------------+--------+--------+---------------+---------+---------+------------------------+------+--------+----------+------------+----------------------------------------------+\r\n\r\n|    1 | SIMPLE      | first  | index  | PRIMARY       | PRIMARY | 96      | NULL                   |  100 | 100.00 |   100.00 |     100.00 | Using index; Using temporary; Using filesort |\r\n\r\n|    1 | SIMPLE      | second | eq_ref | PRIMARY       | PRIMARY | 96      | huff_20170614.first.id |    1 |   1.00 |   100.00 |     100.00 | Using index                                  |\r\n\r\n+------+-------------+--------+--------+---------------+---------+---------+------------------------+------+--------+----------+------------+----------------------------------------------+\r\n\r\nIn our real-life example, the table was ~2GB and tmp_table_size was 64MB, so this meant that the temp table actually went to disk.\r\n\r\nNotable status counter differences:\r\n\r\nHandler_commit\t1\t1\r\n\r\nHandler_read_first\t1\t1\r\n\r\nHandler_read_key\t906\t906\r\n\r\nHandler_read_next\t906\t906\r\n\r\nHandler_read_rnd\t0\t906\r\n\r\nHandler_read_rnd_next\t0\t907\r\n\r\nHandler_tmp_write\t0\t906\r\n","DateCreated":"2019-05-08T22:20:05.6185153+01:00","DateModified":"2019-05-08T22:20:05.6185153+01:00"},"10":{"ID":10,"Priority":1,"Status":5,"Version":"","TargetVersion":"","Summary":"LP:452112 - MariaDB: Debug information not included in .deb packages","Description":"Current deb builds for mariadb do not include debug information. Debug info (at least for the standalone server binary and for embedded server binary) should be included, as otherwise it won't be possible to analyze bugs.\r\n\r\nSee also https://lists.launchpad.net/maria-developers/msg01241.html\r\n","DateCreated":"2019-05-08T22:24:35.4296096+01:00","DateModified":"2019-05-08T22:24:35.4296096+01:00"}},"Tasks":{"1":{"ID":1,"Priority":5,"Status":2,"TargetVersion":"10.4","Summary":"Create MariaDB named commands","Description":"\r\n\r\nMariaDB Server makes use of a lot of commands named mysql something. We should introduce mariadb names for the commands.\r\nManual article examples (compare ease of reading)\r\nALTERNATIVE 1a (change mysql to maria, preserve existing underscore)\r\n\r\nThe mariadump client utility performs logical backups, producing a set of SQL statements. mariadump requires at least the SELECT privilege for dumped tables. mariadump output can include ALTER DATABASE statements that change the database collation. If your tables are primarily MyISAM tables, consider using the mariahotcopy instead, for better performance than mariadump of backup and restore operations. See Section “mariahotcopy — A Database Backup Program”. mariadump can retrieve and dump table contents row by row. There are in general three ways to use mariadump in order to dump a set of one or more tables.\r\n\r\nshell> mariadump [options] db_name [tbl_name ...]\r\n\r\nshell> mariadump [options] --databases db_name ...\r\n\r\nshell> mariadump [options] --all-databases\r\n\r\nmariadump is not to be confused with mariadumpslow, which is a tool to examine the slow query log. There are also mariaaccess and mariaadmin. And you will certainly find mariashow useful, and maybe even mariaslap. maria_secure_installation can be invoked without arguments, so can maria_upgrade.\r\nALTERNATIVE 1b (change mysql to mariadb, preserve existing underscore)\r\n\r\nThe mariadbdump client utility performs logical backups, producing a set of SQL statements. mariadbdump requires at least the SELECT privilege for dumped tables. mariadbdump output can include ALTER DATABASE statements that change the database collation. If your tables are primarily MyISAM tables, consider using the mariadbhotcopy instead, for better performance than mariadbdump of backup and restore operations. See Section “mariadbhotcopy — A Database Backup Program”. mariadbdump can retrieve and dump table contents row by row. There are in general three ways to use mariadbdump in order to dump a set of one or more tables.\r\n\r\nshell> mariadbdump [options] db_name [tbl_name ...]\r\n\r\nshell> mariadbdump [options] --databases db_name ...\r\n\r\nshell> mariadbdump [options] --all-databases\r\n\r\nmariadbdump is not to be confused with mariadbdumpslow, which is a tool to examine the slow query log. There are also mariadbaccess and mariadbadmin. And you will certainly find mariadbshow useful, and maybe even mariadbslap. mariadb_secure_installation can be invoked without arguments, so can mariadb_upgrade.\r\nALTERNATIVE 2a (change mysql to maria, use dash, replace underscore to dash)\r\n\r\nThe maria-dump client utility performs logical backups, producing a set of SQL statements. maria-dump requires at least the SELECT privilege for dumped tables. maria-dump output can include ALTER DATABASE statements that change the database collation. If your tables are primarily MyISAM tables, consider using the maria-hotcopy instead, for better performance than maria-dump of backup and restore operations. See Section “maria-hotcopy — A Database Backup Program”. maria-dump can retrieve and dump table contents row by row. There are in general three ways to use maria-dump in order to dump a set of one or more tables.\r\n\r\nshell> maria-dump [options] dbname [tblname ...]\r\n\r\nshell> maria-dump [options] --databases dbname ...\r\n\r\nshell> maria-dump [options] --all-databases\r\n\r\nmaria-dump is not to be confused with maria-dump-slow, which is a tool to examine the slow query log. There are also maria-access and maria-admin. And you will certainly find maria-show useful, and maybe even maria-slap. maria-secure-installation can be invoked without arguments, so can maria-upgrade.\r\nALTERNATIVE 2b (change mysql to mariadb, use dash, replace underscore to dash)\r\n\r\nThe mariadb-dump client utility performs logical backups, producing a set of SQL statements. mariadb-dump requires at least the SELECT privilege for dumped tables. mariadb-dump output can include ALTER DATABASE statements that change the database collation. If your tables are primarily MyISAM tables, consider using the mariadb-hotcopy instead, for better performance than mariadb-dump of backup and restore operations. See Section “mariadb-hotcopy — A Database Backup Program”. mariadb-dump can retrieve and dump table contents row by row. There are in general three ways to use mariadb-dump in order to dump a set of one or more tables.\r\n\r\nshell> mariadb-dump [options] dbname [tblname ...]\r\n\r\nshell> mariadb-dump [options] --databases dbname ...\r\n\r\nshell> mariadb-dump [options] --all-databases\r\n\r\nmariadb-dump is not to be confused with mariadb-dump-slow, which is a tool to examine the slow query log. There are also mariadb-access and mariadb-admin. And you will certainly find mariadb-show useful, and maybe even mariadb-slap. mariadb-secure-installation can be invoked without arguments, so can mariadb-upgrade.\r\n","DateCreated":"2019-05-08T20:40:45.1924995+01:00","DateModified":"2019-05-08T20:40:45.1924995+01:00"},"2":{"ID":2,"Priority":5,"Status":1,"TargetVersion":"","Summary":"Does mariabackup option \"--databases-exclude\"support Wildcard Character?","Description":"Does mariabackup option \"--databases-exclude\"support Wildcard Character?\r\ne.g:\r\n\r\nmariabackup --backup \r\n\r\n      --target-dir /path/to/backup \\\r\n\r\n      --user user_name --password user_passwd\r\n\r\n      --databases-exclude=\"cache_*\"\r\n","DateCreated":"2019-05-08T20:42:09.5272477+01:00","DateModified":"2019-05-08T20:42:09.5272477+01:00"},"3":{"ID":3,"Priority":2,"Status":1,"TargetVersion":"","Summary":"[Xfce] The browser resets the window size when returning from full-screen mode","Description":"What did you do?\r\n================\r\n1. Maximize the window with browser to the whole desktop (with the help of the window maximizer button)\r\n2. Press F11\r\n3. Press F11 again\r\n\r\nWhat happened?\r\n==============\r\nThe size of the window has changed to a \"default value\". \r\n\r\n\r\nWhat should have happened?\r\n==========================\r\nThe window should be open to the entire desktop.\r\n\r\nIs there anything else we should know?\r\n======================================\r\nI use Xubuntu 16.04 and Firefox Quantum 60.0.2.\r\nThe problem occurs only when the (system) header is disabled.\r\nThe same problem occurs after opening the video in full-screen mode.","DateCreated":"2019-05-08T21:28:03.0603659+01:00","DateModified":"2019-05-08T21:28:03.0603659+01:00"},"4":{"ID":4,"Priority":3,"Status":2,"TargetVersion":"","Summary":"Add EnvironmentFile=/etc/sysconfig/mariadb to MariaDB service's unit file","Description":"It's normal and expected for systemd services to provide a way to set environment variables without direct modification of the servic's until file (that may be lost during upgrade etc). In case of gardb it's done via /etc/sysconfig/garb, in case of MySQL server it's /etc/sysconfig/mysql, see https://dev.mysql.com/doc/refman/8.0/en/using-systemd.htm.\r\n\r\nFor some reason MariaDB service's unit file included with RPMs for CentOS 7 etc does NOT contain any predefined EnvironmentFile setting. I think it should be added, so service reads whatever definitions (like LD_PRELOAD) are present in /etc/sysconfig/mysql (for compatibility) and/or /etc/sysconfig/mariadb without the need to change unit file or do extra \"systemctl set-environment ...\" calls,\r\n","DateCreated":"2019-05-08T22:22:05.3166888+01:00","DateModified":"2019-05-08T22:22:05.3166888+01:00"},"5":{"ID":5,"Priority":1,"Status":3,"TargetVersion":"","Summary":"Document deprecation/disappearance of innodb-file-io-threads","Description":"It looks like innodb-file-io-threads option/variable no longer exists, but the KB still mentions it as an active one: https://mariadb.com/kb/en/library/mysqld-options/#-innodb-file-io-threads","DateCreated":"2019-05-08T22:26:18.551969+01:00","DateModified":"2019-05-08T22:26:18.551969+01:00"}},"ImportResult":null,"ExportResult":null}