- developers - lists.mariadb.org

Re: 264aef71a37: Have ha_partition ignore HA_EXTRA..CHILDREN extra() calls if no myisamrg
by Sergei Golubchik 23 Feb '24

23 Feb '24

Hi, Michael, On Feb 23, Michael Widenius wrote: > revision-id: 264aef71a37 (mariadb-10.6.17-15-g264aef71a37) > parent(s): 042c3fc432b > author: Michael Widenius > committer: Michael Widenius > timestamp: 2024-02-22 09:20:51 +0200 > message: > > Have ha_partition ignore HA_EXTRA..CHILDREN extra() calls if no myisamrg > > diff --git a/sql/ha_partition.cc b/sql/ha_partition.cc > index e174ef8d219..f2b3ab3aff9 100644 > --- a/sql/ha_partition.cc > +++ b/sql/… [View More]

1 0

Re: Second part of review of XA patches in bb-10.6-MDEV-31949
by Kristian Nielsen 17 Feb '24

17 Feb '24

Andrei Elkin <andrei.elkin(a)mariadb.com> writes: > > I think it's better to define a proper API for this. > > It feels an overkill in the current case. "Define a proper API" means that you think through how this should work for different storage engines and TC_LOG implementations. And then you document those thoughts. I don't understand how doing this can be overkill, you are re-defining how durability of external XA transactions are coordinated between the server layer … [View More]and the storage engines. >> And vice versa, for every xid >> X_COMPLETE found, roll it back if X is found prepared in InnoDB, > > Actually >> else commit it. > safer must be to commit both X_COMPLETE and X. Generally even if > X_COMPLETE and X commits in this order, X can race to be durable before > crash. Right, and "commit X" would mean "XA ROLLBACK X" if the "complete" operation is a rollback. I didn't follow why rolling back the X_COMPLETE would be a problem, but it's not important for now; I can see how committing should work as well. >> One reason I'm asking is I'm wondering how this will affect the scheduling >> of XA in parallel replication? Like, in case of duplicate XID, the X_PREPARE >> or X_COMPLETE could conflict between two transactions, so maybe the >> wakeup_subsequent_commits can only be done after the XA COMMIT of the >> X_PREPARE and X_COMMIT? Which I think would require changes to the current >> patch. > > Hmm, right now in the branch (see Sliding_window etc) for the duplicate > XID case, like > P_1,C_2,P_3,C_4 > P - XA-PREPARE, C - ... commit) P_3 can start only when C_2 wakes it up. Yes. So my point is: In this case, the C_2 cannot do wakeup_subsequent_commits inside the group commit along with run_commit_ordered(), that would be too early. C_2 must first complete the extra internal XA transaction on the mysql.gtid_slave_pos table before it can wakeup P_3. Or is this not a problem? > My bad I left out a mention of the plan, to treat - call it > XAC_0 -> XAP_1 > dependency - by the optimistic retry. That's in order to minimize the Aha. So if we optimistically apply the XA PREPARE, and we get an error ER_XAER_DUPID that the XID already exists as prepared, then we would do a retry like other temporary errors, is that what you mean? Is this implemented in the current patch? Maybe it is automatically handled by convert_kill_to_deadlock_error() ? If this XAC_0 -> XAP_1 conflict can be handled by normal optimistic rollback/retry, then I agree there is no need to handle the dependency explicitly in the sliding window. That was not clear to me from the code. >> This I don't understand at all. >> How can an XA PREPARE/COMMIT/ROLLBACK be marked as "ddl" ? > > That's a binlogging whim. E.g a temp table "DDL" involved into > transaction. I was not sure I could quickly eliminate the flagging, so > decided to live with it. Aha, right, CREATE TEMPORARY TABLE can be part of a transaction and thus DDL (in statement-based binlogging). I missed that. > But I could not get the meaning of > > ... handler.cc to avoid special cases for whether the binlog is enabled > The burden of checking whether the binlog is enabled, e.g for taking a > TC_LOG->log_and_order() path is on the programmer. If binlog is globally enabled (--log-bin option), then tc_log is an object of class MYSQL_BIN_LOG. Else it is class TC_LOG_MMAP (or TC_LOG_DUMMY). So a virtual function can be used instead of a conditional on opt_bin_log. But binlogging can be disabled on a transaction-by-transaction basis by setting SQL_LOG_BIN, is that the case here? Then yes, it seems an explicit condition will be needed. - Kristian. [View Less]

1 0

[Andrei Elkin] Re: Third/last part of review of XA patches in bb-10.6-MDEV-31949
by andrei.elkin＠pp.inet.fi 12 Feb '24

12 Feb '24

Kristian, Using the chance to forwarding the reply that did not make to the developers list, let me comment on binlog_commit() that I might hastily rate as a big subject. > Hi Andrei, > > Here is the last part of my review of the changes in branch > bb-10.6-MDEV-31949: > >> --- a/sql/log_event_server.cc >> +++ b/sql/log_event_server.cc >> @@ -3324,6 +3324,11 @@ Gtid_log_event::Gtid_log_event(THD *thd_arg, uint64 seq_no_arg, >> … [View More] flags2|= thd->lex->sql_command == SQLCOM_XA_PREPARE ? >> FL_PREPARED_XA : FL_COMPLETED_XA; >> xid.set(xid_state.get_xid()); >> + >> + // multi-engine external completion is unaware of the prepared engine #. >> + extra_engines= thd->transaction->all.ha_list ? >> + ha_count_rw_2pc(thd_arg, >> + thd_arg->in_multi_stmt_transaction_mode()) - 1 : 0; > > I cannot understand the comment. > I assume it's about that an XA COMMIT of a detached xid (different > session > that the XA PREPARE, or replication slave) - right? Correct. > > But what is the consequences of not knowing extra_engines in this > case? > Is the extra_engines used for anything than the optional truncate of > the > binlog after master crash? In the multi-engine case when the # of prepared in Engine is greater than in the value in GTID event then the trx is partially committed, so is decided for to be completed. > >> @@ -4175,7 +4173,8 @@ int XA_prepare_log_event::do_commit() >> thd->lex->xid= &xid; >> if (!one_phase) >> { >> - if ((res= thd->wait_for_prior_commit())) >> + if (thd->is_current_stmt_binlog_disabled() && >> + (res= thd->wait_for_prior_commit())) > > Why?000 > Is it to allow XA_PREPARE_LOG_EVENT to group commit with other event > groups? > That makes sense, but then shouldn't the wait_for_prior_commit() be > somewhere in handler.cc (or maybe xa.cc)? > Look at Xid_log_event::do_commit(), it doesn't have any > wait_for_prior_commit(), that is in ha_commit_one_phase(). Right. > (The wait-for-prior-commit facility is designed as a server-level > feature, > not a replication-level, it could in principle be used outside of > replication, even though it not currently (IIRC).) I think I understood what you mean which is follow the ha_commit_one_phase() pattern. [todo] > >> diff --git a/sql/sql_array.h b/sql/sql_array.h >> index 8610e971016..94c6d86ac05 100644 >> --- a/sql/sql_array.h >> +++ b/sql/sql_array.h >> @@ -172,6 +172,8 @@ template <class Elem> class Dynamic_array >> >> size_t size() const { return array.elements; } >> >> + size_t max_size() const { return array.max_element; } >> + > > Not that it hurts that much, but I think you can avoid adding this. Left over of a cleanup from the previous version. [todo] > >> --- a/sql/log.cc >> +++ b/sql/log.cc >> @@ -85,6 +85,7 @@ const char *log_bin_index= 0; >> const char *log_bin_basename= 0; >> >> MYSQL_BIN_LOG mysql_bin_log(&sync_binlog_period); >> +XID_cache_element *xid_cache_search(THD *thd, XID *xid); > > This duplicates the same declaration in sql/xa.h > So probably this can be removed now? [todo] > >> @@ -2186,16 +2202,13 @@ int binlog_commit(THD *thd, bool all, bool >> ro_1pc) >> } >> >> if (cache_mngr->trx_cache.empty() && >> - (thd->transaction->xid_state.get_state_code() != XA_PREPARED || >> - !(thd->ha_data[binlog_hton->slot].ha_info[1].is_started() && >> - thd->ha_data[binlog_hton->slot].ha_info[1].is_trx_read_write()))) >> + (!(is_preparing_xa(thd) || is_prepared_xa(thd)) || >> + !is_binlog_read_write(thd))) > > This binlog_commit function started out as the "commit" method for the > binlog handlerton. But that is not how it is used anymore, now it is > called > directly from several different places, including some places specific > to > external user XA. > > A large part of the function has become complex conditions like this > that > are basically impossible for the developer (or reviewer) to understand > exactly why this set of conditions is the correct one. Again, the > function > is trying to guess what the caller wants, depending on the state of > various > thd properties etc. > > Instead, the function should be split into multiple functions. One > that is > called to handle XA PREPARE; one for binlog_commit_by_xid(), etc. Then > each > version should be able to greatly reduce the number of conditions > needed. > > And if any logic is shared between the different versions, this can > then be > put in smaller functions that are called from each version. [my initial response was: Your proposal leads to a significant refactoring, which I add only one hunk. I'd really consider that as a separate exercise I think I've understood. Sure, binlog_prepare() that logs XA-PREPARE can call a common function with binlog_commit() that originally intended to log only normal trx:s. I can look into that. [todo] >> >> @@ -11050,10 +11124,30 @@ bool >> Recovery_context::decide_or_assess(xid_recovery_member *member, int >> round, >> } >> else if (member->in_engine_prepare < last_gtid_engines) >> { >> - DBUG_ASSERT(member->in_engine_prepare > 0); >> + DBUG_ASSERT(member->in_engine_prepare > 0 || >> + /* XA event over xid that was completed in engines */ >> + (member->xid == 0 && >> + member->xa_binlog_state >= xid_recovery_member::XA_PREPARE)); >> + if (member->in_engine_prepare > 0 && >> + member->xa_binlog_state == xid_recovery_member::XA_PREPARE) >> + { > > The function decide_or_assess() is hard to understand. > I can see that it's called in 3 cases: XID_LOG_EVENT, XA PREPARE, XA > COMMIT/ROLLBACK. > So again here, I think this should be split into 3 different > functions, one > for each case (or maybe two, one for internal XID and one for external > XA, > whatever makes the more sense). > > I don't understand the comment "XA event over xid that was completed > in > engines". But I remember seeing something in log_event_server.cc that > sets > extra_engines to 0 in the external user XA case. > > So is that the case here, that last_gtid_engines==0 marks this is user > XA? > That would be much clearer if the logic for internal and external XA > was not > mixed in the same function. > >> + /* >> + the first of the two options below can occur when round 1 turns to 2 >> + so event sequences across binlog files are like the following >> + g_current(B_0, xid), ... g_last(B^*, xid), >> + where `g`:s are xa event groups associated with the same member via >> + common `xid`. As elsewhere B_0 is the 2nd round's first and, >> + B^* the 1st round binlogs. >> + */ >> if (reset_truncate_coord(member->binlog_coord < last_gtid_coord ? >> member->binlog_coord.second : pos)) > > I'm trying to understand what is going on here. I think this is refering to > the way binlog files are scanned? > First we scan the most recent binlog file. At the end of it, we may find > that the latest binlog checkpoint is in an earlier file, then we continue to > scan those older files. So for example we may scan in order: > > master-bin.000004, master-bin.000002, master-bin.000003 > > This means that we may see the XA PREPARE and XA COMMIT out-of-order as we > go from the first scan to the remaining - "round 1 turns to 2". > > But I don't understand that this is "the first of the two options below": > > member->binlog_coord < last_gtid_coord ? member->binlog_coord.second : pos > > If last_gtid_coord is in master-bin.000002, but member->binlog_coord is in > master-bin.000004 ("round 1 turns to 2"), then I would think we would have > > member->binlog_coord > last_gtid_coord > > So it would be the *second* of the options below? What am I missing here? > > Also, I would suggest a simplification here. After we do the initial scan in > round 1, if we find that we need to scan more files, then reset all the > state and hashes. And then do a full scan of all necessary binlog files > in-order. In my example, that would be: > > master-bin.000004 > <reset state> > master-bin.000002 > master-bin.000003 > master-bin.000004 > > This way, the logic becomes simpler, as we don't need logic to handle seeing > events out-of-order. > > This will mean that occasionally we will need to scan one more binlog file, > but I think that is a good trade-off. The recovery code is notoriously > difficult to test and get correct, because it's an unusual occurence to need > recovery in the first place, and very hard to reproduce any problems. And > robust and correct recovery is the whole point of having XA in the first > palce. So making the code as simple as possible is crucial. > > The extra cost is small: > > - For one, recovery is a rare event > - Re-scanning only costs CPU, the file is already read from storage into OS > buffers. > - Normally, the binlog checkpoint completes shortly after rotation, so the > last binlog file that must be rescanned will be short. > > So simply redo the scan from the correct point when a "round 2" is needed, > to avoid any logic to handle seeing events out-of-order during recovery. > > >> @@ -11182,7 +11311,9 @@ void Recovery_context::process_gtid(int round, Gtid_log_event *gev, >> last_gtid_no2pc= false; >> last_gtid_standalone= >> (gev->flags2 & Gtid_log_event::FL_STANDALONE) ? true : false; >> - if (do_truncate && last_gtid_standalone) >> + /* a standalone XA-COMPLETE is truncation-safe */ >> + if (do_truncate && last_gtid_standalone >> + && !(gev->flags2 & Gtid_log_event::FL_COMPLETED_XA)) >> update_binlog_unsafe_coord_if_needed(linfo); > > I don't understand the comment here. > What is "a standalone XA-COMPLETE" ? > What does it mean to be "truncation-safe"? > Please reword the comment to be comprehensible. > >> + /* >> + the user XA events are decided similary to the normal transaction, >> + difference is just in the timing which is Gtid event read now. >> + */ > > I don't understand this comment. Please reformulate to make it > understandable. What does normal transactions do, and what does user > XA do? > How does the timing differ, and why? > >> + sql_print_warning("GTID %u-%u-%s XA transaction state % " >> + "in binlog file:%s pos:%llu is inconsistent " >> + "with the former same state; " >> + "consider filtered binary loggig.", >> + ctx.last_gtid.domain_id, ctx.last_gtid.server_id, >> + buf, is_xac ? "'COMPLETE'" : "'PREPARE'", >> + linfo->log_file_name, ctx.last_gtid_coord.second); > > Typo: s/loggig/logging/ > > Also, the phrase "the former same state" is not understandable (at > least to > me), what does it mean? Please reformulate. > I also have to be short now in replying to the rest, which I'll expand on tomorrow morning. Now just this. >> @@ -909,14 +910,16 @@ ha_commit_checkpoint_request(void *cookie, void (*pre_hook)(void *)) >> there's no need to rollback here as all transactions must >> be rolled back already >> */ >> -void ha_close_connection(THD* thd) >> +void ha_close_connection(THD* thd, bool skip_binlog) > > I tried to understand what the purpose of this change is. > ha_close_connection() is now called from two new places related to > replication of XA PREPARE. And in one place with this mysterious > skip_binlog=true condition. > > My guess is that this is to detach the XA PREPARE from the current > THD? The reason is described in 6cbb63f592 commit name II.1. ``` OTHER CHANGES, to ha_close_connection() are done to keep the binlog hton connection alive for the end of XA-prepare's binlogging on the parallel slave. Other than the binlog hton:s are disconnected from the slave transaction handler in TC_LOG::run_commit_ordered() which makes the prepared transaction engine objects available to binlog-commit-order followers. ``` In other words XAP_k in run_commit_ordered() releases with ha_close_connection(thd, true) all resources that the following XAC_k+i need. The latter does not need XAP_k to release binlog which it still need while being is this stack of TC_LOG::run_commit_ordered MYSQL_BIN_LOG::trx_group_commit_leader MYSQL_BIN_LOG::write_transaction_to_binlog_events MYSQL_BIN_LOG::write_transaction_to_binlog binlog_flush_cache binlog_commit_flush_xa_prepare binlog_commit binlog_prepare prepare_or_error ha_prepare > So > that it becomes visible in XA RECOVER to other threads, and (maybe > more > important) to make the THD available to run another transaction? One > of the > things InnoDB does in close_connection is: trx_disconnect_prepared(). > > But I think the purpose of the close_connection handlerton call is > different And you must have meant 'ha_close_connection', right? > - it is to free things at the end of a client connection? Well, with the ha- prefix, it's to disconnect THD from all transaction branches/htons. > So it can do > more > than detach the current external XA - which is presumably the reason > for the > need to hack it to not call the specific binlog_close_connection()? > > So this shouldn't hijack the close_connection to mean > "detach_xa_trx". It > looks like we need an explicit handlerton call for detaching an xa > transaction, that can then be used instead of this hack in > ha_close_connection(). I need to digest that in the morning. Generally I don't see a problem to conduct hton release flexibly. > >> -static my_bool xacommit_handlerton(THD *unused1, plugin_ref plugin, >> - void *arg) >> +my_bool xacommit_handlerton(THD *unused1, plugin_ref plugin, void *arg) >> { >> handlerton *hton= plugin_hton(plugin); >> - if (hton->recover) >> + if (hton->recover && hton != binlog_hton) >> { >> hton->commit_by_xid(hton, ((struct xahton_st *)arg)->xid); > >> -int ha_commit_or_rollback_by_xid(XID *xid, bool commit) >> +int ha_commit_or_rollback_by_xid(XID *xid, bool commit, THD *thd) >> { > >> + if (!skip_binlog) >> + { >> + // when binlog is ON start from its transaction branch >> if (commit) >> binlog_commit_by_xid(binlog_hton, xid); >> else >> binlog_rollback_by_xid(binlog_hton, xid); > > I think the purpose if this is to ensure that commit/rollback of XA > calls > into the binlog code first (to binlog before doing it in the engines). > > But as I think Serg also remarked already, this should not be > needed. The > binlog is the (internal) transaction coordinator when it is enabled, > we have > a dedicated interface for that, class TC_LOG. This is used elsewhere > in > handler.cc to avoid special cases for whether the binlog is enabled or > not. > > So it looks like there needs to be extra virtual functions in TC_LOG > for > the XA commit/rollback, and these can then be called on the tc_log > object > and do what is required. I need to process that tomorrow. > Then some logic in handler.cc can be omitted/simplified. > > ---- > > I think with this, I got through all of the code changes in branch > bb-10.6-MDEV-31949 as of commit > 1bdc9bc8077c79e55cb4dbfc7cc47a193d17d54f > from January 20, 2024. So this concludes the current review. I have > not > looked at the test cases in this round of review. Thanks again! Cheers, Andrei [View Less]

1 1

Re2: Starting on review of XA patches in bb-10.6-MDEV-31949
by andrei.elkin＠pp.inet.fi 12 Feb '24

12 Feb '24

Salve Kristian, this is the 2nd follow-up replay to your 1st review, to cover few deferred notes in particular. I mark with [x] any addressed [todo] note. > Kristian Nielsen <knielsen(a)knielsen-hq.org> writes: > >> Hi Andrei, > So you're in this context > > 1 innobase_commit_by_xid( > 2 /*===================*/ > 3 handlerton* hton, > 4 XID* xid) /*!< in: X/Open XA transaction > identification */ > 5 { > 6 … [View More]DBUG_ASSERT(hton == innodb_hton_ptr); > 7 > 8 DBUG_EXECUTE_IF("innobase_xa_fail", > 9 return XAER_RMFAIL;); > 10 > 11 if (high_level_read_only) { > 12 return(XAER_RMFAIL); > 13 } > 14 > 15 if (trx_t* trx = trx_get_trx_by_xid(xid)) { > 16 THD *thd = current_thd; > 17 if (thd) > 18 thd_binlog_pos(thd, &trx->mysql_log_file_name, > 19 &trx->mysql_log_offset); > 20 if (trx->mysql_log_offset > 0) > 21 trx->flush_log_later = true; > > and actually considering `trx->mysql_log_offset` as the return value? > ... I could not grasp your way of thinking in this place. > Could you please expand on? In another mail yesterday concerning with this part I suggested to replace if (trx->mysql_log_offset > 0) with if (thd_binlog_format(thd) != UNSPEC) > > >> >> It would be much better if this could be done similar to how normal >> commits >> are done. We have the commit_ordered handlerton which does the fast >> part of >> commit-to-memory. And then the slow part which does the rest in commit >> handlerton. > > Notice that the *slow* part of the XA commit is the same as in the > normal > case: > > trx_commit_complete_for_mysql > *innobase_commit* > commit_one_phase_2 > ha_commit_one_phase > > The fast parts of the two - the XA and normal - could be unified - into > > innobase_commit_ordered() > > but I did not think (had time for) towards that. > >> >> So we could have commit_by_xid_ordered() and commit_by_xid(). The >> problem is >> that commit_by_xid_ordered() needs to release the xid, so it becomes >> harder >> to get hold of the transaction (which the commit handlerton just finds >> from >> the thd). Maybe commit_by_xid_ordered() would need to return some trx >> handle. >> >> What part of the commit is skipped by this in InnoDB? Is it _only_ >> flushing >> the log up to the LSN of the commit record, or are there more things? > > > So the xa's fast innobase_commit_by_xid() skips a "slow" > trx_commit_complete_for_mysql() > and the same does the normal's innobase_commit_ordered(). > >> >> How is another storage engine supposed to handle this? This also needs >> to be >> documented somewhere, eg. in comments on the handlerton calls. >> >> This also means that the --innodb-flush-log-at-trx-commit=3 is not >> supported >> for XA COMMIT when it's done from a different session that the XA >> PREPARE. >> For example, with --innodb-flush-log-at-trx-commit=3 but >> --sync-binlog=0, >> normal transactions are still durable in InnoDB. But external XA >> transactions will not be. This could be solved if the commit_by_xid() >> was >> split into a fast and slow part similar to normal commit. > > [todo] Let me process that carefully later. [x] My 1st reply already said, just missed to underscore that, that the XA also the slow and fast commit paths, therefore iflatc=3 is good. The past paths are commit_one_phase_2 and innobase_commit_by_xid functionally equivalent. One would vote to unify them, and I'd accept that. > >> >> Probably you'll say that this works for now, and perhaps that's true. >> But at >> least it really _really_ needs some detailed comments explaining what >> is >> going on, and what should be kept in mind if modifying/improving this >> code >> in the future, otherwise it will be impossible for the next developer >> to >> understand. > > [todo] Agreed. > >> >> >>> diff --git a/sql/log.cc b/sql/log.cc >>> index e54d4087d46..7e3dbeaef29 100644 >>> --- a/sql/log.cc >>> +++ b/sql/log.cc >> >>> @@ -9365,18 +9381,84 @@ TC_LOG::run_commit_ordered(THD *thd, bool >>> all) >>> all ? thd->transaction->all.ha_list : >>> thd->transaction->stmt.ha_list; >>> >>> mysql_mutex_assert_owner(&LOCK_commit_ordered); >>> - for (; ha_info; ha_info= ha_info->next()) >>> + >>> + if (!ha_info) >> >> Your patch puts a lot of XA/binlog logic into run_commit_ordered(). >> But >> that's not appropriate. The run_commit_ordered() function has the >> purpose >> solely to call the commit_ordered() handlerton in each engine, and >> it's also >> called from TC_LOG_MMAP (that's why it's TC_LOG::, not >> MYSQL_BIN_LOG::). >> >> Instead make a new function that handles the part of binlog commit >> that >> needs to be done under LOCK_commit_ordered. Be sure to comment it that >> this >> runs under this highly contended mutex, and needs to do as little work >> as >> possible. Agreed. [todo] >> >> This function can then call run_commit_ordered(). It can also be used >> to >> share a few bits of code that are now duplicated between the >> opt_optimize_thread_scheduling and !opt_optimize_thread_scheduling >> cases, >> like: >> >> ++num_commits; >> run_commit_ordered(...) >> entry->thd->wakeup_subsequent_commits() >> if (entry->queued_by_other) { ... } >> >> (if it makes sense to do so in the actual code, only if makes things >> simpler, that's easier to decide when working with the actual code). Will see. [todo] >> >> There sesems to be 3 new cases: >> >> - XA PREPARE >> - XA COMMIT from same session as XA PREPARE >> - XA COMMIT of detached XA transaction >> >> The logic to decide this was already done earlier, when the event >> group was >> queued for group commit, eg. to decide which end_event to use. For >> example >> there is no need to call xid_cache_search() here, that can be done >> earlier >> (and it's also wrong to call my_error() here, as we're in the wrong >> thread >> context). Indeed. The error's thread context was overlooked. Thanks for catching it. >> >> So the logic about what to do should be placed higher up in the call >> stack, >> ideally where we already know, eg. binlog_commit_by_xid(). And errors >> such >> as ER_XAER_NOTA can be thrown already there, and the decision on what >> to do >> put consisely as a flag into the struct group_commit_entry, so the >> logic >> during binlog group commit becomes as simple as possible. >> >> (It might even make sense to simply store a function pointer to the >> method >> that should be called to handle this; normal, attached XA COMMIT, >> detached >> XA COMMIT, XA PREPARE. But again it's easier when actually working >> with the >> code to decide if this is simpler than a couple flags or case on enum >> value). > > > [todo] Let me process the above suggestion block a bit later. [x] To the need of xid_cache_search() etc, let me combine this part to reply to the 2nd review mail's notes around a Dummy XID though ... (read on) > >> >>> + if (thd->rgi_slave && thd->rgi_slave->is_parallel_exec) >>> + { >>> + DBUG_ASSERT(thd->transaction->xid_state.is_dummy_XA()); >>> + >>> + auto xs= xid_cache_search(thd, xid); >>> + if(!xs) >>> + { >>> + my_error(ER_XAER_NOTA, MYF(0)); >>> + return; >>> + } >>> + else >>> + { >>> + thd->transaction->xid_state.xid_cache_element= xs; >>> + } >>> + DBUG_ASSERT(thd->transaction->xid_state.is_recovered()); >>> + } >>> + >>> plugin_foreach(NULL, commit ? xacommit_handlerton : >>> xarollback_handlerton, >>> MYSQL_STORAGE_ENGINE_PLUGIN, &xaop); >>> xid_cache_delete(thd); >> >> Why this check for thd->rgi_slave->is_parallel_exec here deep in >> binlog >> group commit? That seems very wrong, and again no comment whatsoever >> explaining what is going on, so again I have to try to guess... >> >> I'm guessing that it's something with calling xid_cache_delete() which >> needs >> to find thd->transaction->xid_state.xid_cache_element ? > > Right. > >> >> But this could be done much better in a higher-level part of the code >> that's >> specific to xa and (parallel) replication, not here deep down in >> binlog >> group commit code. >> >> And why should this be needed only for parallel replication, not for >> non-parellel replication or non-replication SQL? I don't understand. > > > It's a kind of optimization for the slave side, in that `xid` of XAP > (Prepare) gets released for a XAC (Commit) the soonest. > ... the reason's said here. >> >> >> A couple remarks about the tests I pushed to >> knielsen_mdev31949_review: >> > >> >> I think the failures in rpl.rpl_recovery_xa I linked are from >> duplicate XID >> in XA PREPARE ; XA COMMIT ; XA PREPARE. There comes an ambiguity >> between >> whether the transaction prepared in the engine at recovery is from the >> first >> XA PREPARE (so it should be committed) or the second (so it should be >> rolled >> back). This happens when the second XA PREPARE is in the engine but >> not >> binlogged. >> >> One way to fix could actually be to use the binlog position stored >> inside >> InnoDB. That's a neat reminder on the Innodb durably committed binlog position! Indeed with the duplicate xid case there exists this trouble of identification and the Engine is able and must come to rescue. >> If the XA COMMIT is before this binlog position, then we know the >> first transaction became durable in InnoDB, so we know we have to roll >> back >> the current prepared transaction in the engine. Otherwise we should >> commit >> it. I see the recovery code already uses binlog positions to decide >> the fate >> of pending XA transactions in the engines, but I did not yet fully >> understand what is going on there. >> >> Another way could be to binlog an extra event before the second XA >> PREPARE >> with the duplicate XID. Well, this does not cover the case of prior to crash XA PREPARE only made to Engine. The above consulting with Innodb resolves the ambiguity generally. >> This event would mean "This xid is now durable >> committed in the engine", so we would know not to commit any prepared >> trx in >> the engine with the same xid. This event would only be needed in case >> of XA >> PREPARE with a duplicate xid. This though requires the prior XA COMMIT >> to >> fsync inside InnoDB first, I think. > > Thanks for this piece of analysis! > [todo] I am processing this a bit later. [x] > >> >> >> Another test failure in rpl.slave_provision_xa I think happens because >> the >> XA PREPARE does not update the binlog position inside innodb. >> Maybe the way >> to solve this problem is to have the slave be idempotent, and silently >> ignore any duplicate XA PREPARE if the xid is already prepared? Indeed, the fact that XAP does not update the innodb's binlog position must cause few concerns. I still need to have a close look. [todo] > > [todo] I am processing this a bit later. [x] > >> >> >> About the "ToDo" of using some Xid_list_log_event. I think this is a >> way >> during recovery to get the state of the XA PREPARE/COMMIT's in earlier >> binlog files before the last binlog checkpoint, if I understood things >> correctly? > > True. > >> >> Another way to do this is to delay the binlog checkpoints until all XA >> PREPARE in a binlog file have been XA COMMIT'ted (or XA ROLLBACK'ed), >> so >> that they will all be scanned during recovery. I actually implemented >> this >> in my branch knielsen_mdev32020. The advantage of this is that we may >> anyway >> need this to provide the XA PREPARE events to a slave set up from a >> logical >> backup (like mysqldump), to avoid purging XA PREPARE events too early. >> And >> then it's a bit simpler to simply scan an extra file rather than both >> scan >> and read Xid_list_log_event. But I think Xid_list_log_event could work >> fine >> as well, just mentioning it for you to consider. > > Xlle can't be generally avoided so we have to take conservatively > first. > To add up to a separate thread discussion over this problematics, I've grown more inclined with the first implementation basing on your idea to defer Binlog-CheckPoint. That might work for a majority of XA users, Xid list to be added later if time permits or will be convincingly :-) requested. Now I am turning to account the rest - 2,3 reviews you sent that only glanced through - to reply within few hours. Cheers, Andrei [View Less]

2 2

Re: 64-bit time_t support in MariaDB?
by Otto Kekäläinen 12 Feb '24

12 Feb '24

> it's not related to time, it's just an assert that's verified at compile > time, that is, during compilation. Like static_assert in C++. > > I just need to find 32-bit arm somewhere. Unfortunately I don't have any with SSH access to give. Seems the reported of MDEV-33429 has armhf hardware and is helping out. Once you have a patch, I can push it to Launchpad builders (or you can do it yourself, anyone with an Launchpad can push code to their own PPA). Related, Cfarm is a … [View More]

2 1

Re2: thd_binlog_pos / Re: Starting on review of XA patches in bb-10.6-MDEV-31949
by andrei.elkin＠pp.inet.fi 08 Feb '24

08 Feb '24

>> So you're in this context >> >> 1 innobase_commit_by_xid( > >> 18 thd_binlog_pos(thd, &trx->mysql_log_file_name, >> 19 &trx->mysql_log_offset); >> 20 if (trx->mysql_log_offset > 0) >> 21 trx->flush_log_later = true; >> >> and actually considering `trx->mysql_log_offset` as the return value? >> Could you please expand on? > > My point was … [View More]

1 0

Starting on review of XA patches in bb-10.6-MDEV-31949
by Kristian Nielsen 07 Feb '24

07 Feb '24

Hi Andrei, It's not an easy review of the XA patches in bb-10.6-MDEV-31949. One thing that would have helped is a clear design document that lists all the different concerns and points that need to be considered around recovery, backup, replication, possible conflicts, etc. This could then be referenced by the reviewer to see if a given concern was considered, and if so what the resolution is. But I'll try as best I can with what we have. As I'm trying to understand the different issues, I … [View More]found making test cases to check my understanding helped. I have pushed some to the branch knielsen_mdev31949_review . They seem to show some problems in the current patch: mysql-test/suite/rpl/t/rpl_recovery_xa.test - A trx is left in "XA PREPARED" state after crash recovery, even though the binlogged event group is incomplete (it should have been rolled back). - A trx is committed after crash recovery, even though it is not binlogged (it should have been rolled back). - An incomplete XA event group is output by mysqlbinlog with a "ROLLBACK" at the end which gives an error (should probably be XA ROLLBACK). mysql-test/suite/rpl/t/rpl_recovery_xa2.test - An XA PREPAREd transaction is lost after recovery when it was binlogged in an earlier binlog file. (I think this might be the issue that's referenced as a ToDo with Xid_list_log_event in patch?) mysql-test/suite/mariabackup/slave_provision_xa.test - An XA PREPARE does not update the binlog position inside InnoDB. So a slave provisioned with mariabackup from this starting position can fail with duplicate apply of XA PREPARE. Some other initial remarks follow. Since there's a desire to progress with this quickly, I thought it would be useful to send these early, so you can start looking/discussions before I complete building an understanding of the whole design and code picture. > diff --git a/storage/innobase/handler/ha_innodb.cc b/storage/innobase/handler/ha_innodb.cc > index 17e512bf34e..a4cdc72e18c 100644 > --- a/storage/innobase/handler/ha_innodb.cc > +++ b/storage/innobase/handler/ha_innodb.cc > @@ -17144,8 +17144,16 @@ innobase_commit_by_xid( > } > > if (trx_t* trx = trx_get_trx_by_xid(xid)) { > + THD *thd = current_thd; > + if (thd) > + thd_binlog_pos(thd, &trx->mysql_log_file_name, > + &trx->mysql_log_offset); I guess this is necessary to get the correct binlog position stored inside innodb to match the binlogged XA COMMIT event, right? But will it work if called during crash recovery? During crash recovery, we could actually update the position, but only if this is trx is the last event group in the old binlog. Maybe that's better handled elsewhere (there are currently other cases where the binlog position in InnoDB can be wrong immediately after restart until the first new transaction is committed). > + if (trx->mysql_log_offset > 0) > + trx->flush_log_later = true; But I'm guessing this is the motivation. If called from within binlog group commit, you want to disable fsync in InnoDB. But if called from recovery, or when binlog is disabled, this condition will be false, and fsync will be done. If so, that's a _really_ obfuscated way to do it :-/ It surely needs a very clear comment explaining what is going on. And it is very fragile. What if thd_binlog_pos() is changed at some point, to slightly change the conditions under which it returns 0 or not? That could silently break this code. It would be much better if this could be done similar to how normal commits are done. We have the commit_ordered handlerton which does the fast part of commit-to-memory. And then the slow part which does the rest in commit handlerton. So we could have commit_by_xid_ordered() and commit_by_xid(). The problem is that commit_by_xid_ordered() needs to release the xid, so it becomes harder to get hold of the transaction (which the commit handlerton just finds from the thd). Maybe commit_by_xid_ordered() would need to return some trx handle. What part of the commit is skipped by this in InnoDB? Is it _only_ flushing the log up to the LSN of the commit record, or are there more things? How is another storage engine supposed to handle this? This also needs to be documented somewhere, eg. in comments on the handlerton calls. This also means that the --innodb-flush-log-at-trx-commit=3 is not supported for XA COMMIT when it's done from a different session that the XA PREPARE. For example, with --innodb-flush-log-at-trx-commit=3 but --sync-binlog=0, normal transactions are still durable in InnoDB. But external XA transactions will not be. This could be solved if the commit_by_xid() was split into a fast and slow part similar to normal commit. Probably you'll say that this works for now, and perhaps that's true. But at least it really _really_ needs some detailed comments explaining what is going on, and what should be kept in mind if modifying/improving this code in the future, otherwise it will be impossible for the next developer to understand. > diff --git a/sql/log.cc b/sql/log.cc > index e54d4087d46..7e3dbeaef29 100644 > --- a/sql/log.cc > +++ b/sql/log.cc > @@ -9365,18 +9381,84 @@ TC_LOG::run_commit_ordered(THD *thd, bool all) > all ? thd->transaction->all.ha_list : thd->transaction->stmt.ha_list; > > mysql_mutex_assert_owner(&LOCK_commit_ordered); > - for (; ha_info; ha_info= ha_info->next()) > + > + if (!ha_info) Your patch puts a lot of XA/binlog logic into run_commit_ordered(). But that's not appropriate. The run_commit_ordered() function has the purpose solely to call the commit_ordered() handlerton in each engine, and it's also called from TC_LOG_MMAP (that's why it's TC_LOG::, not MYSQL_BIN_LOG::). Instead make a new function that handles the part of binlog commit that needs to be done under LOCK_commit_ordered. Be sure to comment it that this runs under this highly contended mutex, and needs to do as little work as possible. This function can then call run_commit_ordered(). It can also be used to share a few bits of code that are now duplicated between the opt_optimize_thread_scheduling and !opt_optimize_thread_scheduling cases, like: ++num_commits; run_commit_ordered(...) entry->thd->wakeup_subsequent_commits() if (entry->queued_by_other) { ... } (if it makes sense to do so in the actual code, only if makes things simpler, that's easier to decide when working with the actual code). There sesems to be 3 new cases: - XA PREPARE - XA COMMIT from same session as XA PREPARE - XA COMMIT of detached XA transaction The logic to decide this was already done earlier, when the event group was queued for group commit, eg. to decide which end_event to use. For example there is no need to call xid_cache_search() here, that can be done earlier (and it's also wrong to call my_error() here, as we're in the wrong thread context). So the logic about what to do should be placed higher up in the call stack, ideally where we already know, eg. binlog_commit_by_xid(). And errors such as ER_XAER_NOTA can be thrown already there, and the decision on what to do put consisely as a flag into the struct group_commit_entry, so the logic during binlog group commit becomes as simple as possible. (It might even make sense to simply store a function pointer to the method that should be called to handle this; normal, attached XA COMMIT, detached XA COMMIT, XA PREPARE. But again it's easier when actually working with the code to decide if this is simpler than a couple flags or case on enum value). > + if (thd->rgi_slave && thd->rgi_slave->is_parallel_exec) > + { > + DBUG_ASSERT(thd->transaction->xid_state.is_dummy_XA()); > + > + auto xs= xid_cache_search(thd, xid); > + if(!xs) > + { > + my_error(ER_XAER_NOTA, MYF(0)); > + return; > + } > + else > + { > + thd->transaction->xid_state.xid_cache_element= xs; > + } > + DBUG_ASSERT(thd->transaction->xid_state.is_recovered()); > + } > + > plugin_foreach(NULL, commit ? xacommit_handlerton : xarollback_handlerton, > MYSQL_STORAGE_ENGINE_PLUGIN, &xaop); > xid_cache_delete(thd); Why this check for thd->rgi_slave->is_parallel_exec here deep in binlog group commit? That seems very wrong, and again no comment whatsoever explaining what is going on, so again I have to try to guess... I'm guessing that it's something with calling xid_cache_delete() which needs to find thd->transaction->xid_state.xid_cache_element ? But this could be done much better in a higher-level part of the code that's specific to xa and (parallel) replication, not here deep down in binlog group commit code. And why should this be needed only for parallel replication, not for non-parellel replication or non-replication SQL? I don't understand. A couple remarks about the tests I pushed to knielsen_mdev31949_review: I think the failures in rpl.rpl_recovery_xa I linked are from duplicate XID in XA PREPARE ; XA COMMIT ; XA PREPARE. There comes an ambiguity between whether the transaction prepared in the engine at recovery is from the first XA PREPARE (so it should be committed) or the second (so it should be rolled back). This happens when the second XA PREPARE is in the engine but not binlogged. One way to fix could actually be to use the binlog position stored inside InnoDB. If the XA COMMIT is before this binlog position, then we know the first transaction became durable in InnoDB, so we know we have to roll back the current prepared transaction in the engine. Otherwise we should commit it. I see the recovery code already uses binlog positions to decide the fate of pending XA transactions in the engines, but I did not yet fully understand what is going on there. Another way could be to binlog an extra event before the second XA PREPARE with the duplicate XID. This event would mean "This xid is now durable committed in the engine", so we would know not to commit any prepared trx in the engine with the same xid. This event would only be needed in case of XA PREPARE with a duplicate xid. This though requires the prior XA COMMIT to fsync inside InnoDB first, I think. Another test failure in rpl.slave_provision_xa I think happens because the XA PREPARE does not update the binlog position inside innodb. Maybe the way to solve this problem is to have the slave be idempotent, and silently ignore any duplicate XA PREPARE if the xid is already prepared? About the "ToDo" of using some Xid_list_log_event. I think this is a way during recovery to get the state of the XA PREPARE/COMMIT's in earlier binlog files before the last binlog checkpoint, if I understood things correctly? Another way to do this is to delay the binlog checkpoints until all XA PREPARE in a binlog file have been XA COMMIT'ted (or XA ROLLBACK'ed), so that they will all be scanned during recovery. I actually implemented this in my branch knielsen_mdev32020. The advantage of this is that we may anyway need this to provide the XA PREPARE events to a slave set up from a logical backup (like mysqldump), to avoid purging XA PREPARE events too early. And then it's a bit simpler to simply scan an extra file rather than both scan and read Xid_list_log_event. But I think Xid_list_log_event could work fine as well, just mentioning it for you to consider. That's it for now, I will continue trying to understand what is going on and complete the review. - Kristian. [View Less]

1 2

Re: Xlle question
by andrei.elkin＠pp.inet.fi 06 Feb '24

06 Feb '24

>> Xlle can't be generally avoided so we have to take conservatively >> first. > > Why it cannot be avoided? What is it needed for? The main purpose is to discern at recovery a prepared XA (XAP) that was never binlogged from one that was. Your case >>> Another way to do this is to delay the binlog checkpoints until all >>> XA >>> PREPARE in a binlog file have been XA COMMIT'ted assumes that XAC must always come in "normal" time. But of … [View More]

1 0

Fwd/Re: Starting on review of XA patches in bb-10.6-MDEV-31949
by andrei.elkin＠pp.inet.fi 05 Feb '24

05 Feb '24

From: Andrei Elkin <andrei.elkin(a)mariadb.com> Subject: Re: Starting on review of XA patches in bb-10.6-MDEV-31949 To: Kristian Nielsen <knielsen(a)knielsen-hq.org> Cc: developers(a)lists.mariadb.org, andrei elkin <andrei.elkin(a)mariadb.com>, Michael Widenius <michael.widenius(a)gmail.com> Date: Mon, 05 Feb 2024 14:11:01 +0200 (36 minutes, 34 seconds ago) Organization: MariaDB Kristian Nielsen <knielsen(a)knielsen-hq.org> writes: > Hi Andrei, > > … [View More]It's not an easy review of the XA patches in bb-10.6-MDEV-31949. > > One thing that would have helped is a clear design document that lists > all > the different concerns and points that need to be considered around > recovery, backup, replication, possible conflicts, etc. This could then > be > referenced by the reviewer to see if a given concern was considered, > and if > so what the resolution is. I am sorry for discomfort the lack of the above caused. Unfortunately the IV recovery part has consumed more time than was expected (which is at the final stage of my local testing). > But I'll try as best I can with what we > have. If you'll need any online discussion to compensate for that I'd be glad to do it. > > As I'm trying to understand the different issues, I found making test > cases > to check my understanding helped. I have pushed some to the branch > knielsen_mdev31949_review. Thank you for streamlining it this way! > They seem to show some problems in the current > patch: > > mysql-test/suite/rpl/t/rpl_recovery_xa.test > - A trx is left in "XA PREPARED" state after crash recovery, even > though > the binlogged event group is incomplete (it should have been > rolled > back). > - A trx is committed after crash recovery, even though it is not > binlogged (it should have been rolled back). > - An incomplete XA event group is output by mysqlbinlog with a > "ROLLBACK" > at the end which gives an error (should probably be XA ROLLBACK). > > mysql-test/suite/rpl/t/rpl_recovery_xa2.test > - An XA PREPAREd transaction is lost after recovery when it was > binlogged > in an earlier binlog file. (I think this might be the issue that's > referenced as a ToDo with Xid_list_log_event in patch?) > > mysql-test/suite/mariabackup/slave_provision_xa.test > - An XA PREPARE does not update the binlog position inside InnoDB. > So a > slave provisioned with mariabackup from this starting position can > fail > with duplicate apply of XA PREPARE. > > Some other initial remarks follow. Since there's a desire to progress > with > this quickly, I thought it would be useful to send these early, so you > can > start looking/discussions before I complete building an understanding > of the > whole design and code picture. Super! > > >> diff --git a/storage/innobase/handler/ha_innodb.cc >> b/storage/innobase/handler/ha_innodb.cc >> index 17e512bf34e..a4cdc72e18c 100644 >> --- a/storage/innobase/handler/ha_innodb.cc >> +++ b/storage/innobase/handler/ha_innodb.cc >> @@ -17144,8 +17144,16 @@ innobase_commit_by_xid( >> } >> >> if (trx_t* trx = trx_get_trx_by_xid(xid)) { >> + THD *thd = current_thd; >> + if (thd) >> + thd_binlog_pos(thd, &trx->mysql_log_file_name, >> + &trx->mysql_log_offset); > > I guess this is necessary to get the correct binlog position stored > inside > innodb to match the binlogged XA COMMIT event, right? Correct. > But will it work if called during crash recovery? Right. The unpushed part IV makes sure of that. > > During crash recovery, we could actually update the position, but only > if > this is trx is the last event group in the old binlog. Maybe that's > better > handled elsewhere (there are currently other cases where the binlog > position > in InnoDB can be wrong immediately after restart until the first new > transaction is committed). > >> + if (trx->mysql_log_offset > 0) >> + trx->flush_log_later = true; > > But I'm guessing this is the motivation. If called from within binlog > group > commit, you want to disable fsync in InnoDB. But if called from > recovery, or > when binlog is disabled, this condition will be false, and fsync will > be > done. Indeed. > > If so, that's a _really_ obfuscated way to do it :-/ I remember it was a very quick decision for myself to miss this sort of reflection. [todo] I'll add comments of course. > It surely needs a very clear comment explaining what is going on. > And it is very fragile. What if thd_binlog_pos() is changed at some > point, > to slightly change the conditions under which it returns 0 or not? That > could silently break this code. So you're in this context 1 innobase_commit_by_xid( 2 /*===================*/ 3 handlerton* hton, 4 XID* xid) /*!< in: X/Open XA transaction identification */ 5 { 6 DBUG_ASSERT(hton == innodb_hton_ptr); 7 8 DBUG_EXECUTE_IF("innobase_xa_fail", 9 return XAER_RMFAIL;); 10 11 if (high_level_read_only) { 12 return(XAER_RMFAIL); 13 } 14 15 if (trx_t* trx = trx_get_trx_by_xid(xid)) { 16 THD *thd = current_thd; 17 if (thd) 18 thd_binlog_pos(thd, &trx->mysql_log_file_name, 19 &trx->mysql_log_offset); 20 if (trx->mysql_log_offset > 0) 21 trx->flush_log_later = true; and actually considering `trx->mysql_log_offset` as the return value? ... I could not grasp your way of thinking in this place. Could you please expand on? > > It would be much better if this could be done similar to how normal > commits > are done. We have the commit_ordered handlerton which does the fast > part of > commit-to-memory. And then the slow part which does the rest in commit > handlerton. Notice that the *slow* part of the XA commit is the same as in the normal case: trx_commit_complete_for_mysql *innobase_commit* commit_one_phase_2 ha_commit_one_phase The fast parts of the two - the XA and normal - could be unified - into innobase_commit_ordered() but I did not think (had time for) towards that. > > So we could have commit_by_xid_ordered() and commit_by_xid(). The > problem is > that commit_by_xid_ordered() needs to release the xid, so it becomes > harder > to get hold of the transaction (which the commit handlerton just finds > from > the thd). Maybe commit_by_xid_ordered() would need to return some trx > handle. > > What part of the commit is skipped by this in InnoDB? Is it _only_ > flushing > the log up to the LSN of the commit record, or are there more things? So the xa's fast innobase_commit_by_xid() skips a "slow" trx_commit_complete_for_mysql() and the same does the normal's innobase_commit_ordered(). > > How is another storage engine supposed to handle this? This also needs > to be > documented somewhere, eg. in comments on the handlerton calls. > > This also means that the --innodb-flush-log-at-trx-commit=3 is not > supported > for XA COMMIT when it's done from a different session that the XA > PREPARE. > For example, with --innodb-flush-log-at-trx-commit=3 but > --sync-binlog=0, > normal transactions are still durable in InnoDB. But external XA > transactions will not be. This could be solved if the commit_by_xid() > was > split into a fast and slow part similar to normal commit. [todo] Let me process that carefully later. > > Probably you'll say that this works for now, and perhaps that's true. > But at > least it really _really_ needs some detailed comments explaining what > is > going on, and what should be kept in mind if modifying/improving this > code > in the future, otherwise it will be impossible for the next developer > to > understand. [todo] Agreed. > > >> diff --git a/sql/log.cc b/sql/log.cc >> index e54d4087d46..7e3dbeaef29 100644 >> --- a/sql/log.cc >> +++ b/sql/log.cc > >> @@ -9365,18 +9381,84 @@ TC_LOG::run_commit_ordered(THD *thd, bool all) >> all ? thd->transaction->all.ha_list : >> thd->transaction->stmt.ha_list; >> >> mysql_mutex_assert_owner(&LOCK_commit_ordered); >> - for (; ha_info; ha_info= ha_info->next()) >> + >> + if (!ha_info) > > Your patch puts a lot of XA/binlog logic into run_commit_ordered(). But > that's not appropriate. The run_commit_ordered() function has the > purpose > solely to call the commit_ordered() handlerton in each engine, and it's > also > called from TC_LOG_MMAP (that's why it's TC_LOG::, not > MYSQL_BIN_LOG::). > > Instead make a new function that handles the part of binlog commit that > needs to be done under LOCK_commit_ordered. Be sure to comment it that > this > runs under this highly contended mutex, and needs to do as little work > as > possible. > > This function can then call run_commit_ordered(). It can also be used > to > share a few bits of code that are now duplicated between the > opt_optimize_thread_scheduling and !opt_optimize_thread_scheduling > cases, > like: > > ++num_commits; > run_commit_ordered(...) > entry->thd->wakeup_subsequent_commits() > if (entry->queued_by_other) { ... } > > (if it makes sense to do so in the actual code, only if makes things > simpler, that's easier to decide when working with the actual code). > > There sesems to be 3 new cases: > > - XA PREPARE > - XA COMMIT from same session as XA PREPARE > - XA COMMIT of detached XA transaction > > The logic to decide this was already done earlier, when the event group > was > queued for group commit, eg. to decide which end_event to use. For > example > there is no need to call xid_cache_search() here, that can be done > earlier > (and it's also wrong to call my_error() here, as we're in the wrong > thread > context). > > So the logic about what to do should be placed higher up in the call > stack, > ideally where we already know, eg. binlog_commit_by_xid(). And errors > such > as ER_XAER_NOTA can be thrown already there, and the decision on what > to do > put consisely as a flag into the struct group_commit_entry, so the > logic > during binlog group commit becomes as simple as possible. > > (It might even make sense to simply store a function pointer to the > method > that should be called to handle this; normal, attached XA COMMIT, > detached > XA COMMIT, XA PREPARE. But again it's easier when actually working with > the > code to decide if this is simpler than a couple flags or case on enum > value). [todo] Let me process the above suggestion block a bit later. > >> + if (thd->rgi_slave && thd->rgi_slave->is_parallel_exec) >> + { >> + DBUG_ASSERT(thd->transaction->xid_state.is_dummy_XA()); >> + >> + auto xs= xid_cache_search(thd, xid); >> + if(!xs) >> + { >> + my_error(ER_XAER_NOTA, MYF(0)); >> + return; >> + } >> + else >> + { >> + thd->transaction->xid_state.xid_cache_element= xs; >> + } >> + DBUG_ASSERT(thd->transaction->xid_state.is_recovered()); >> + } >> + >> plugin_foreach(NULL, commit ? xacommit_handlerton : >> xarollback_handlerton, >> MYSQL_STORAGE_ENGINE_PLUGIN, &xaop); >> xid_cache_delete(thd); > > Why this check for thd->rgi_slave->is_parallel_exec here deep in binlog > group commit? That seems very wrong, and again no comment whatsoever > explaining what is going on, so again I have to try to guess... > > I'm guessing that it's something with calling xid_cache_delete() which > needs > to find thd->transaction->xid_state.xid_cache_element ? Right. > > But this could be done much better in a higher-level part of the code > that's > specific to xa and (parallel) replication, not here deep down in binlog > group commit code. > > And why should this be needed only for parallel replication, not for > non-parellel replication or non-replication SQL? I don't understand. It's a kind of optimization for the slave side, in that `xid` of XAP (Prepare) gets released for a XAC (Commit) the soonest. > > > A couple remarks about the tests I pushed to knielsen_mdev31949_review: > > > I think the failures in rpl.rpl_recovery_xa I linked are from duplicate > XID > in XA PREPARE ; XA COMMIT ; XA PREPARE. There comes an ambiguity > between > whether the transaction prepared in the engine at recovery is from the > first > XA PREPARE (so it should be committed) or the second (so it should be > rolled > back). This happens when the second XA PREPARE is in the engine but not > binlogged. > > One way to fix could actually be to use the binlog position stored > inside > InnoDB. If the XA COMMIT is before this binlog position, then we know > the > first transaction became durable in InnoDB, so we know we have to roll > back > the current prepared transaction in the engine. Otherwise we should > commit > it. I see the recovery code already uses binlog positions to decide the > fate > of pending XA transactions in the engines, but I did not yet fully > understand what is going on there. > > Another way could be to binlog an extra event before the second XA > PREPARE > with the duplicate XID. This event would mean "This xid is now durable > committed in the engine", so we would know not to commit any prepared > trx in > the engine with the same xid. This event would only be needed in case > of XA > PREPARE with a duplicate xid. This though requires the prior XA COMMIT > to > fsync inside InnoDB first, I think. Thanks for this piece of analysis! [todo] I am processing this a bit later. > > > Another test failure in rpl.slave_provision_xa I think happens because > the > XA PREPARE does not update the binlog position inside innodb. Maybe the > way > to solve this problem is to have the slave be idempotent, and silently > ignore any duplicate XA PREPARE if the xid is already prepared? [todo] I am processing this a bit later. > > > About the "ToDo" of using some Xid_list_log_event. I think this is a > way > during recovery to get the state of the XA PREPARE/COMMIT's in earlier > binlog files before the last binlog checkpoint, if I understood things > correctly? True. > > Another way to do this is to delay the binlog checkpoints until all XA > PREPARE in a binlog file have been XA COMMIT'ted (or XA ROLLBACK'ed), > so > that they will all be scanned during recovery. I actually implemented > this > in my branch knielsen_mdev32020. The advantage of this is that we may > anyway > need this to provide the XA PREPARE events to a slave set up from a > logical > backup (like mysqldump), to avoid purging XA PREPARE events too early. > And > then it's a bit simpler to simply scan an extra file rather than both > scan > and read Xid_list_log_event. But I think Xid_list_log_event could work > fine > as well, just mentioning it for you to consider. Xlle can't be generally avoided so we have to take conservatively first. > > > That's it for now, I will continue trying to understand what is going > on and > complete the review. Thank you for so prompt piece of feedback! I'll return with a reply within a work day to [todo] tagged. Cheers, Andrei ---------- [View Less]

2 1

Review of MDEV-7850: Extend GTID Binlog Events with Thread Id
by Kristian Nielsen 31 Jan '24

31 Jan '24

> commit c37b2087b4abe576f1b0391c8d379dba6299dcb5 (origin/bb-11.4-MDEV-7850) > Author: Brandon Nesterenko <brandon.nesterenko(a)mariadb.com> > Date: Mon Jul 10 18:53:19 2023 +0300 > > MDEV-7850: Extend GTID Binlog Events with Thread Id > > This patch augments Gtid_log_event with the user thread-id. > In particular that compensates for the loss of this info in > Rows_log_events. Here's my review of this patch (which was already pushed, … [View More]but there are some misunderstandings that should be fixed; Monty asked me about it as he had been told that I had reviewed the patch, but that is not the case (until now)): > diff --git a/sql/log_event.cc b/sql/log_event.cc > index 41ddc038594..f167284ad1d 100644 > --- a/sql/log_event.cc > +++ b/sql/log_event.cc > @@ -1901,11 +1904,12 @@ Query_log_event::begin_event(String *packet, ulong ev_offset, > /* > Currently we only need to replace GTID event. > The length of GTID differs depending on whether it contains commit id. > + And/or thread id. > */ > - DBUG_ASSERT(data_len == LOG_EVENT_HEADER_LEN + GTID_HEADER_LEN || > - data_len == LOG_EVENT_HEADER_LEN + GTID_HEADER_LEN + 2); > - if (data_len != LOG_EVENT_HEADER_LEN + GTID_HEADER_LEN && > - data_len != LOG_EVENT_HEADER_LEN + GTID_HEADER_LEN + 2) > + DBUG_ASSERT(data_len >= LOG_EVENT_HEADER_LEN + GTID_HEADER_LEN && > + data_len <= LOG_EVENT_HEADER_LEN + GTID_HEADER_LEN + 2 + 9); > + if (data_len < LOG_EVENT_HEADER_LEN + GTID_HEADER_LEN || > + data_len > LOG_EVENT_HEADER_LEN + GTID_HEADER_LEN + 2 + 9) This change of the asserts/checks doesn't make any sense. They were there originally, because we are hand-crafting a BEGIN query event, and there were only two cases of the length, so two fixed-sized events were crafted by two different code paths. And the asserts are there to safe-guard if the length would change. But now this check was already wrong, as more optional info has been added to the GTID event without updating the code for Query_log_event::begin_event() (XID, flags_extra, extra_engines). But the new code in this patch is generic and can handle any length GTID event, so these checks no longer make any sense, and should just be removed (rather than attempt to correct them). > @@ -1926,9 +1930,19 @@ Query_log_event::begin_event(String *packet, ulong ev_offset, > } > else > { > - DBUG_ASSERT(data_len == LOG_EVENT_HEADER_LEN + GTID_HEADER_LEN + 2); > - /* Put in an empty time_zone_str to take up the extra 2 bytes. */ > - int2store(q + Q_STATUS_VARS_LEN_OFFSET, 2); > + DBUG_ASSERT(data_len <= LOG_EVENT_HEADER_LEN + GTID_HEADER_LEN + 11); This assert should not be there. The length can be larger than 11, and the code below works fine for arbitrary length. > + DBUG_ASSERT(data_len >= LOG_EVENT_HEADER_LEN + GTID_HEADER_LEN + 2); > + > + /* Put in an empty time_zone_str to take up the extra 2 plus the number of > + dummy_bytes. */ > + size_t dummy_bytes= > + data_len - (LOG_EVENT_HEADER_LEN + GTID_HEADER_LEN + 2); > + > + int2store(q + Q_STATUS_VARS_LEN_OFFSET, dummy_bytes + 2); > + for (size_t i= 0; i < dummy_bytes; i++) > + q[Q_DATA_OFFSET + i]= Q_DUMMY; > + q+= dummy_bytes; > + > q[Q_DATA_OFFSET]= Q_TIME_ZONE_CODE; > q[Q_DATA_OFFSET+1]= 0; /* Zero length for empty time_zone_str */ > q[Q_DATA_OFFSET+2]= 0; /* Zero terminator for empty db */ This new code with Q_DUMMY is generic and independent of the length of the GTID event. That's good! But it needs a comment explaining (that the Q_DUMMY is unknown to the old slave, which will make the old slave just skip the rest of the user vars, which are just there to fill up the extra bytes). And now the code that adds the dummy empty time_zone_str is redundant, and should just be removed, as should the code for the case of data_len == LOG_EVENT_HEADER_LEN + GTID_HEADER_LEN. Since now any extra bytes in the GTID can just be replaced by Q_DUMMY. This way, the code will be fixed to work for all the possible GTID lengths, and be shorter and simpler (and a lot less confusing). > @@ -1584,6 +1584,9 @@ Query_log_event::Query_log_event(const uchar *buf, uint event_len, > sa_seq_no = uint8korr(pos); > pos+= 8; > } > + } > + case Q_DUMMY: > + { > break; > } > default: > /* That's why you must write status vars in growing order of code */ > DBUG_PRINT("info",("Query_log_event has unknown status vars (first has code: %u), skipping the rest of them", (uint) *(pos-1))); > pos= (const uchar*) end; // Break loop There's no point in having this code. Q_DUMMY is supposed to be higher than any known status var type, so it will be ignored anyway in the `default` case. And besides, server that knows about Q_DUMMY can never receive it in the first place. (But if this is really wanted, then don't delete the `break` statement in the prior switch case, that's likely to introduce a bug the next time a case is added. And assign Q_DUMMY the next free value, not the value 255). > diff --git a/sql/log_event.h b/sql/log_event.h > index 473506fafed..d324642ad5e 100644 > --- a/sql/log_event.h > +++ b/sql/log_event.h > @@ -324,6 +324,7 @@ class String; > > #define Q_HRNOW 128 > #define Q_XID 129 > +#define Q_DUMMY 255 This needs a good comment, explaining how Q_DUMMY is a code that will be unknown to any server version, and can be used to make the slave skip this (and any following) user vars. diff --git a/mysql-test/suite/rpl/t/rpl_gtid_thread_id.test b/mysql-test/suite/rpl/t/rpl_gtid_thread_id.test new file mode 100644 index 00000000000..b8ed1865da9 --- /dev/null +++ b/mysql-test/suite/rpl/t/rpl_gtid_thread_id.test @@ -0,0 +1,125 @@ +# +# Verify that GTID log events are written into the binary log along with the +# id of the thread which executed the transaction. On the primary, this is the <snip> Disclaimer: I did not review this testcase. - Kristian. [View Less]

2 3