- I am not sure why --rewrite-db is called a "variable". It's an
option.
- The TokuDB build version cannot be just "5.6.16-64.0", because
it will conflict with the regular release having the same
version. IIRC the TokuDB build scripts added something to the
version string, so let's add the same thing here and use it
uniformly.
- TokuDB compression has to be documented with examples, listing
of allowed row_type values, probably on its own page.
- The TokuDB alpha features perhaps should be listed somewhere or
users won't even know about their existence. Perhaps a short page
for alpha features only which links to TokuDB pages on secondary
clustering keys, prescans, NOAR, ONLINE OPTIMIZE, autoincrement
would suffice.
- 250: "have no differences and can be installed side by side if
desired." It needs documenting what will happen on existing
Percona-provided libmysqlclient package upgrade, and please run
the resulting paragraph by BAlexey/raghu/tomislav for feedback.
- 259: "due to a regression in the buffer pool mutex split, a
combination of InnoDB compression, write workload, and multiple
active purge threads could lead to a server crash".
- 261: "Several fixes for multiple XA engine support in the server
have been ported from MariaDB."
- 265: "Fix for bug 1227581, a buffer pool mutex split regression,
was not complete, thus a combination of write workload to InnoDB
compressed table and a tablespace drop could crash the server".
- 269: "The InnoDB page cleaner thread could have incorrectly
decided whether the server is busy or idle on some of its
iterations and consequently issue a too big flush list flushing
request on a busy server, causing performance instabilities"
- 273-275: join: "InnoDB did not handle the cases of asynchronous
and synchronous I/O requests completing partially or being
interrupted."
- 277: "Fixed the upstream bug 70768: the fix for upstream bug
70768 may cause a high rate of RW lock creations and
desctructions, resulting in a performance regression on some
workloads"
- 279: "Debian and Ubuntu packaging has been reworked to meet the
packaging standards"
- 281: add 1274411 there too
- 1277383 needs a RN entry
- 1159625/1199534 may or may not need RN entries
- 1204871 needs a RN entry
- The multiple XA fixes need more details what was fixed in one
bug and in the other.
- 1229598 needs a RN entry
- 1005787 needs a RN entry
- 1264952 needs a RN entry
- 1276445 and 1014477 might get a RN entry, "Percona Server source
tree has been reorganized to match the MySQL source tree layout
closer" "Permissions on some testsuite files have been
corrected"
- 1052636 needs a RN entry
- 338: "availble"
- 363-369: unify "INSTALL PLUGIN" case
- I am not sure why --rewrite-db is called a "variable". It's an provided libmysqlclient package upgrade, and please run raghu/tomislav for feedback.
option.
- The TokuDB build version cannot be just "5.6.16-64.0", because
it will conflict with the regular release having the same
version. IIRC the TokuDB build scripts added something to the
version string, so let's add the same thing here and use it
uniformly.
- TokuDB compression has to be documented with examples, listing
of allowed row_type values, probably on its own page.
- The TokuDB alpha features perhaps should be listed somewhere or
users won't even know about their existence. Perhaps a short page
for alpha features only which links to TokuDB pages on secondary
clustering keys, prescans, NOAR, ONLINE OPTIMIZE, autoincrement
would suffice.
- 250: "have no differences and can be installed side by side if
desired." It needs documenting what will happen on existing
Percona-
the resulting paragraph by BAlexey/
- 259: "due to a regression in the buffer pool mutex split, a
combination of InnoDB compression, write workload, and multiple
active purge threads could lead to a server crash".
- 261: "Several fixes for multiple XA engine support in the server
have been ported from MariaDB."
- 265: "Fix for bug 1227581, a buffer pool mutex split regression,
was not complete, thus a combination of write workload to InnoDB
compressed table and a tablespace drop could crash the server".
- 269: "The InnoDB page cleaner thread could have incorrectly
decided whether the server is busy or idle on some of its
iterations and consequently issue a too big flush list flushing
request on a busy server, causing performance instabilities"
- 273-275: join: "InnoDB did not handle the cases of asynchronous
and synchronous I/O requests completing partially or being
interrupted."
- 277: "Fixed the upstream bug 70768: the fix for upstream bug
70768 may cause a high rate of RW lock creations and
desctructions, resulting in a performance regression on some
workloads"
- 279: "Debian and Ubuntu packaging has been reworked to meet the
packaging standards"
- 281: add 1274411 there too
- 1277383 needs a RN entry
- 1159625/1199534 may or may not need RN entries
- 1204871 needs a RN entry
- The multiple XA fixes need more details what was fixed in one
bug and in the other.
- 1229598 needs a RN entry
- 1005787 needs a RN entry
- 1264952 needs a RN entry
- 1276445 and 1014477 might get a RN entry, "Percona Server source
tree has been reorganized to match the MySQL source tree layout
closer" "Permissions on some testsuite files have been
corrected"
- 1052636 needs a RN entry
- 338: "availble"
- 363-369: unify "INSTALL PLUGIN" case