Commit Graph

78 Commits

Author SHA1 Message Date
Shlomi Noach
3955a6d67f hibernate on critical-load 2017-05-24 08:32:13 +03:00
Shlomi Noach
7517238bd3 Merge branch 'master' into fix-infinite-cutover-loop 2017-03-06 21:50:26 +02:00
Shlomi Noach
06c909bd10 Validating table name length 2017-02-21 17:34:49 -07:00
Shlomi Noach
289ce46a2b Merge branch 'master' into fix-infinite-cutover-loop 2017-02-16 14:16:32 +02:00
Shlomi Noach
ed5dd7f049 Collecting and presenting MySQL versions of applier and inspector 2017-02-07 09:41:33 +02:00
Shlomi Noach
6fa32f6d54 cut-over failure on test-on-replica starts replication again 2017-02-07 09:31:52 +02:00
Shlomi Noach
8d0faa55e3 explicit rollback in ApplyDMLEventQueries() 2017-01-04 08:44:04 +02:00
Shlomi Noach
445c67635d batching DML writes, configurable --dml-batch-size 2017-01-03 14:31:19 +02:00
Shlomi Noach
3e28f462d8 Initial support for batching multiple DMLs when writing to ghost table 2017-01-03 13:44:52 +02:00
rj03hou
ffbd35e180 fix TableEngine correlates to the 3rd placeholder in the template string, not the 1st 2016-12-02 11:56:29 +08:00
Shlomi Noach
9b068ec222 Merge branch 'master' into myisam-gtid 2016-12-01 09:43:38 +01:00
rj03hou
a11bec1785 If the original table is MyISAM and the default engine is Innodb, and the gtid mode is on, there will be error when execute 'LOCK TABLES tbl WRITE, tbl_magic WRITE'. If make the magic cut-over table's engine same with the original table, there will not have this problem. 2016-12-01 16:04:04 +08:00
Shlomi Noach
b00cae11fa retry cut-over 2016-11-17 17:10:17 +01:00
Shlomi Noach
7fa5e405d4 avoid writing heartbeat when throttle commanded by user
when throttling on user command there really is no need for injecting heartbeat. The user commanded, therefore gh-ost complies and trusts the reasoning for throttling. What this will allow is complete quiet time. This, in turn, will allow such features as relocating via orchestrator/pseudo-gtid at time of throttling
2016-10-27 14:51:38 +02:00
Shlomi Noach
ac6159791d merged master, resolved conflicts 2016-10-26 09:57:59 +02:00
Shlomi Noach
bf92eec214 validating table structure on applier and migrator
- reading column list on applier
- comparing original table on applier and migrator, expecting exact column list
- or else bailing out
2016-10-20 11:29:30 +02:00
Shlomi Noach
25166e33c7 solving the enum-as-part-of-pk bug 2016-10-19 15:22:29 +02:00
Shlomi Noach
f9c15127cd simplified applier read of timezone 2016-10-14 12:56:43 +02:00
Shlomi Noach
ac304def4d applier always uses UTC 2016-10-13 13:08:02 +02:00
Shlomi Noach
dbf50afbc7 reading time_zone settings for Inspector and Applier separately.
--time-zone overrides both of them, if given
2016-10-11 16:00:26 +02:00
Shlomi Noach
6750959e1a configurable time-zone, native time parsing 2016-10-10 11:39:57 +02:00
Shlomi Noach
0c35a811f7 setting time_zoe='+00:00' on rowcopy 2016-10-08 11:06:27 +02:00
Shlomi Noach
791d963ea0 Character set recognition and manipulation
- Identifying textual characters sets; converting into specific type when applying dml events
- Refactored `ColumnsList`: introducing `Column` type
- Refactored `unsigned` handling, as part of `Column`
- `Column` type supports `convertArg()`: converting value of argument according to column data type
- DB URI attempts `utf8mb4,utf8,latin1` charsets in that order (first one to be recognized wins)
- Local tests filter by pattern
- Local tests append table schema on failure
- Local tests do not have postpone flag file
- Added character set local tests: `utf8`, `utf8mb4`, `latin1`
2016-09-07 14:24:11 +02:00
Shlomi Noach
2afb86b9e4 support for millisecond throttling
- `--max-lag-millis` is at least `100ms`
- `--heartbeat-interval-millis` introduced; defaults `500ms`, can range `100ms` - `1s`
- Control replicas lag calculated asynchronously to throttle test
  - aggressive when `max-lag-millis < 1000` and when `replication-lag-query` is given
2016-08-30 09:41:59 +02:00
Shlomi Noach
c7edd1ef84 Merge branch 'master' into concurrent-rowcount 2016-08-25 09:44:12 +02:00
Shlomi Noach
1773f338c2 keeping track of delta rows
on concurrent count(*) this means we re-apply delta onto new estimate
2016-08-24 12:16:34 +02:00
Shlomi Noach
4c972184a8 Merge branch 'master' into sql-mode-strict 2016-08-24 09:34:00 +02:00
Shlomi Noach
56fd82a824 Merge pull request #174 from Wattpad/test-on-replica-manual-replication-control
outstanding. Thank you!
2016-08-24 09:12:21 +02:00
Paulo Bittencourt
6b21ade6d0 Check for --test-on-replica-skip-replica-stop in cutOver method 2016-08-23 18:34:10 -04:00
Shlomi Noach
8b76d0e75b DML write sets sql_mode to STRICT ALL TABLES 2016-08-23 11:58:52 +02:00
Shlomi Noach
b63cc3e75e fix INSERT DML handling on renamed column 2016-08-22 16:00:15 +02:00
Shlomi Noach
9cf4819a98 Merge branch 'master' into fix-rename
Wish to incorporate important time_zone fix
2016-08-22 11:54:52 +02:00
Shlomi Noach
1376f0af23 fixed UPDATE dml on renamed column 2016-08-22 08:49:27 +02:00
Paulo Bittencourt
2e43718ef3 Add --test-on-replica-skip-replica-stop flag 2016-08-19 17:34:08 -04:00
Shlomi Noach
6d80340e4f setting time_zone on DML apply 2016-08-19 09:06:00 +02:00
Paulo Bittencourt
a62f9e0754 Add --test-on-replica-manual-replication-control flag
This will wait indefinitely for the replication status to change.
This allows us to run test schema changes in RDS without needing
custom RDS commands in gh-ost.
2016-08-18 11:53:25 -04:00
Shlomi Noach
75e0d12302 simplified error logic; fixed incorrect RowsEstimate handling on error 2016-08-18 13:38:23 +02:00
Shlomi Noach
74593ec010 DML write wrapped in transaction
- solving the golang problem: 'sql: converting Exec argument #2's type: uint64 values with high bit set are not supported'
2016-08-18 13:31:53 +02:00
Shlomi Noach
3a0ee9b4a5 clarified commented transactional apply 2016-08-17 10:50:41 +02:00
Shlomi Noach
4c8edf6372 elaborate error message on applying event data: printing out the error, query and args 2016-08-17 06:50:40 +02:00
Shlomi Noach
596dce5993 elaborate output on error in apply dml 2016-08-15 15:23:30 +02:00
Damian Gryski
e02a49449e all: use time.Since() instead of time.Now().Sub
Patch created with:
    gofmt -w -r 'time.Now().Sub(a) -> time.Since(a)' .
2016-08-02 08:38:56 -04:00
Shlomi Noach
ef59a866d8 Removed legacy 'safe cut-over'
Now that we have the atomic cut-over, the former is redundant
2016-07-16 08:12:19 -06:00
Shlomi Noach
8217536898 supporting --cut-over-lock-timeout-seconds 2016-07-08 10:14:58 +02:00
Shlomi Noach
0191b2897d an atomic cut-over implementation, as per issue #82 2016-06-27 11:08:06 +02:00
Shlomi Noach
96e8419a35 Solved cut-over stall; change of table names
- Cutover would stall after `lock tables` wait-timeout due do waiting on a channel that would never be written to. This has been identified, reproduced, fixed, confirmed.
- Change of table names. Heres the story:
  - Because were testing this even while `pt-online-schema-change` is being used in production, the `_tbl_old` naming convention makes for a collision.
  - "old" table name is now `_tbl_del`, "del" standing for "delete"
  - ghost table name is now `_tbl_gho`
  - when issuing `--test-on-replica`, we keep the ghost table around, and were also briefly renaming original table to "old". Well this collides with a potentially existing "old" table on master (one that hasnt been dropped yet).
  `--test-on-replica` uses `_tbl_ght` (ghost-test)
  - similar problem with `--execute-on-replica`, and in this case the table doesnt stick around; calling it `_tbl_ghr` (ghost-replica)
  - changelog table is now `_tbl_ghc` (ghost-changelog)
  - To clarify, I dont want to go down the path of creating "old" tables with 2 or 3 or 4 or 5 or infinite leading underscored. I think this is very confusing and actually not operations friendly. Its OK that the migration will fail saying "hey, you ALREADY have an old table here, why dont you take care of it first", rather than create _yet_another_ `____tbl_old` table. Were always confused on which table it actually is that gets migrated, which is safe to `drop`, etc.
- just after rowcopy completing, just before cutover, during cutover: marking as point in time _of interest_ so as to increase logging frequency.
2016-06-21 12:56:01 +02:00
Shlomi Noach
62b8a897e3 Retries, better visibility, documentation
- Rowcopy time is bounded by copy end-time
- Retries are configurable via `--default-retries` (default: `60`)
- `migrator` notes the hostname
- `applier` and `inspector` note `impliedKey` (`@@hostname` and `@@port`)
- Added lots of code comments
- Adding documentation for "triggerless design"
2016-06-19 17:55:37 +02:00
Shlomi Noach
23cb8ea7e9 Throttling & critical load
- Added `--throttle-query` param (when returns > 0, throttling applies)
- Added `--critical-load`, similar to `--max-load` but implies panic and quit
- Recoded *-load as `LoadMap`
- More info on *-load throttle/panic
- `printStatus()` now gets printing heuristic. Always shows up on interactive `"status"`
- Fixed `change column` (aka rename) handling with quotes
- Removed legacy `mysqlbinlog` parser code
- Added tests
2016-06-18 21:12:07 +02:00
Shlomi Noach
836d0fe119 Supporting column rename
- Parsing `alter` statement to catch `change old_name new_name ...` statements
- Auto deducing renamed columns
- When suspecting renamed columns, requesting explicit `--approve-renamed-columns` or `--skip-renamed-columns`
- updated tests
2016-06-17 08:03:18 +02:00
Shlomi Noach
96bc3804eb test-on-replica stops replication completely 2016-06-14 12:50:07 +02:00