no sql mode please. make it a separate sysvar.

I am very anxious to use system versioning, but I cannot make the changeover until this feature is available. This is due to the fact that I have existing history data that needs to be moved over to the new system. I am glad to see that it is In Review, and hope to see it in the next MariaDB release.

I also interested in the system versioning. If I could get logical backup by mysqldump, it would be great for me.

The ticket for such a feature seems to be blocked by this ticket. There are several PRs for this JIRA issue, while they are not merged. Could anyone let me know how things are going?

If anyone can do this, I see a security risk.
Will this require the SUPER privilege?

f_razzoli, it's just a convenience feature. It doesn't allow to do anything that one wasn't allowed to do before. Say, you want to have a fake history record that tells that a row with the value V existed from timestamp A to timestamp B. You can do it as

set @@timestamp=A;

insert into t1 values (V);

set @@timestamp=B;

delete from t1 where value=V;

Note, that this is only allowed for certain values of @@secure_timestamp variable. For example, with @@secure_timestamp=YES no one can fake history.

In this issue we want to simplify the above to

set @@system_versioning_insert_history=1;

insert t1 values (V, A, B);

This is primarily to be able to use multi-row inserts in mysqldump. This convenience feature won't do any additional privilege checks, it will work for everyone who can change @@timestamp as above, so it'll obey @@secure_timestamp setting.

Please review bb-10.6-midenok

Please review bb-10.11-midenok-MDEV-16546

I've interrupted the review, as I stumbled on a bug, try this test case:

create table t1(y int primary key,

  row_start timestamp(6) as row start,

  row_end timestamp(6) as row end,

  period for system_time (row_start, row_end))

with system versioning;

insert into t1 values (1, NULL, NULL);

--error ER_WARNING_NON_DEFAULT_VALUE_FOR_GENERATED_COLUMN

insert into t1 values (2, '1980-01-01 00:00:00', '1980-01-01 00:00:01');

set @@system_versioning_insert_history= 1;

insert into t1 values (3, '1980-01-01 00:00:00', '1980-01-01 00:00:01');

select * from t1 for system_time all;

drop table t1;

This INSERT uses a different fill_record() function and does not go into Item_field::fix_fields().

Fixed. It should not go into Item_field::fix_fields() because there is no explicit field list (so no Item_field). It uses different fill_record() for implicit field list, that's normal.

I wanted to split THD::vers_insert_history(Field *) into Field::vers_insert_history(THD *) and THD::vers_insert_history(). But they both cannot be inline (or it sort of hard to rule out without inline files) because Field::vers_insert_history(THD *) must call THD::vers_insert_history() but sql_class.h is not included into field.h. serg, I remember about your inline request.

midenok, you haven't fixed the issue yet. You don't have the test case for it. If you do

-insert into t3(z, row_start, row_end) values (5, '1980-01-01 00:00:00', '1980-01-01 00:00:01');

+insert into t3 values (5, '1980-01-01 00:00:00', '1980-01-01 00:00:01');

you'll see that values are ignored

pushed into bb-10.11-MDEV-16546

I have no objections against pushing bb-10.11-MDEV-16546 as of 87fca0525 into 10.11 branch and releasing it with 10.11.1.

Some notes in no particular order (greenman FYI):

• the final implementation allows INSERT, CREATE .. INSERT, and LOAD; does not allow REPLACE, LOAD .. REPLACE, UPDATE, INSERT .. ON DUPLICATE KEY UPDATE
• in addition to sufficient permissions to change the timestamp, the user performing the operation naturally also needs standard grants for inserting into the table;
• the variable is already in the KB, but the description needs to indicate that it only applies to timestamp-based versioning;
• while inserting history into a versioned table with limit-based partitioning, the user must be aware of MDEV-29674, all injected historical rows are inserted into the first partition. It may be fixed in future;
• in general, functionality is mainly aimed for re-loading the real history by the means of mysqldump / restoration. While arbitrary history inserting also works, the effect may be not quite what the user expects to achieve.
• see also notes in MDEV-16029.