Commit was authored by sujatha.sivakumar and committed by Elkin (latter now assigned to issue). Maybe Andrei can give some pointers of what the original commit tried to achieve so it would be easier to debug what the code is doing?
otto, No pointers can be made at this point but I am going to look into shortly. It feels a bug. You'll see more details from commits.
wenhug, thanks for the report! It's not easy to reproduce though. I tried running your SP while kill -9 the server.
Every time there's a trx in prepared state it gets recovered.
First I am wondering if my method is really equivalent. Maybe you could give it a try to prove practically.
Secondly, could you please tell me your server build options and run time options?
Maybe building and running as you did I will have better luck.
Notice you can gain really helpful warnings into the error log with setting log_warning > 2.
See e.g
if (count_in_prepare > 0 && global_system_variables.log_warnings > 2) |
{
|
sql_print_warning("Could not complete %u number of transactions.", |
count_in_prepare);
|
|
Once you'll have more info in the error log I may have some ideas.
If nothing like that will help us, please consider if I could log in your server and examine the recovery
with gdb.
Thank you and Merry Xmas!
wenhug, to your quick fix,
it's equivalent of executing --tc-heuristic-recover=ROLLBACK. That's why it works out.
After closer look into possible causes of non-decided xids by the 2nd ha_recover() time,
there's a
if (round == 1) |
{
|
if (!binlog_checkpoint_found) |
break; |
escape from the for-loop and after it
if (do_xa) |
{
|
if (binlog_checkpoint_found) |
{
|
#ifndef HAVE_REPLICATION
|
if (ha_recover_complete(&xids)) |
#else |
if (ctx.complete(this, xids)) |
#endif
|
won't be attended.
Could you please examine the last binlog file on whether it contains any Binlog_checkpoint_log_event.
When it does not indeed, your next attempt with log_warnings = 3 obviously won't hit
if (count_in_prepare > 0 && global_system_variables.log_warnings > 2) |
{
|
sql_print_warning("Could not complete %u number of transactions.", |
count_in_prepare);
|
return false; // there's later dry run ha_recover() to error out |
}
|
which otherwise (Binlog_checkpoint event is in place) it must do.
Also more warnings from inside ha_recover_complete would also display then xid:s of undecided sticky transactions.
Hope this will be helpful.
Elkin Thank you very much for looking into the issue!
The issue occurred when creating a snapshot of the file system. I couldn't reproduce with `kill -9` either using the same binaries. Also probably there were some other SQL commands ran before the snapshot was taken but I don't know for sure.
I'll do some more tests and see if there's a steady way to reproduce it. Or maybe can share the data file if that's allowed by our policy.
I just tried starting mysqld with --log_warnings=3 as you suggested:
(It's on another instance with the same issue, so the trx id is different to the original one)
2021-12-26 19:39:04 0 [Note] Loaded 'ha_rocksdb.so' with offset 0x1550a0e1b000 |
2021-12-26 19:39:04 0 [Note] RocksDB: 2 column families found |
2021-12-26 19:39:04 0 [Note] RocksDB: Column Families at start: |
2021-12-26 19:39:04 0 [Note] cf=default |
2021-12-26 19:39:04 0 [Note] write_buffer_size=67108864 |
2021-12-26 19:39:04 0 [Note] target_file_size_base=67108864 |
2021-12-26 19:39:04 0 [Note] cf=__system__ |
2021-12-26 19:39:04 0 [Note] write_buffer_size=67108864 |
2021-12-26 19:39:04 0 [Note] target_file_size_base=67108864 |
2021-12-26 19:39:04 0 [Note] RocksDB: Table_store: loaded DDL data for 0 tables |
2021-12-26 19:39:04 0 [Note] RocksDB: global statistics using get_sched_indexer_t indexer |
2021-12-26 19:39:04 0 [Note] MyRocks storage engine plugin has been successfully initialized. |
2021-12-26 19:39:04 0 [Note] InnoDB: Compressed tables use zlib 1.2.11 |
2021-12-26 19:39:04 0 [Note] InnoDB: Number of pools: 1 |
2021-12-26 19:39:04 0 [Note] InnoDB: Using crc32 + pclmulqdq instructions |
2021-12-26 19:39:04 0 [Note] InnoDB: Using Linux native AIO |
2021-12-26 19:39:04 0 [Note] InnoDB: Initializing buffer pool, total size = 2147483648, chunk size = 134217728 |
2021-12-26 19:39:04 0 [Note] InnoDB: Completed initialization of buffer pool |
2021-12-26 19:39:04 0 [Note] InnoDB: Transaction 126555 was in the XA prepared state. |
2021-12-26 19:39:04 0 [Note] InnoDB: 1 transaction(s) which must be rolled back or cleaned up in total 0 row operations to undo |
2021-12-26 19:39:04 0 [Note] InnoDB: Trx id counter is 126556 |
2021-12-26 19:39:04 0 [Note] InnoDB: 128 rollback segments are active. |
2021-12-26 19:39:04 0 [Note] InnoDB: Creating shared tablespace for temporary tables |
2021-12-26 19:39:04 0 [Note] InnoDB: Starting in background the rollback of recovered transactions |
2021-12-26 19:39:04 0 [Note] InnoDB: Rollback of non-prepared transactions completed |
2021-12-26 19:39:04 0 [Note] InnoDB: Setting file '/rdsdbdata/db/innodb/ibtmp1' size to 12 MB. Physically writing the file full; Please wait ... |
2021-12-26 19:39:04 0 [Note] InnoDB: File '/rdsdbdata/db/innodb/ibtmp1' size is now 12 MB. |
2021-12-26 19:39:04 0 [Note] InnoDB: 10.6.5 started; log sequence number 15159285; transaction id 126557 |
2021-12-26 19:39:04 0 [Note] InnoDB: Loading buffer pool(s) from /rdsdbdata/db/innodb/ib_buffer_pool |
2021-12-26 19:39:04 0 [Note] Recovering after a crash using /rdsdbdata/log/binlog/mysql-bin-changelog |
2021-12-26 19:39:04 0 [Note] Starting table crash recovery... |
2021-12-26 19:39:04 0 [Note] InnoDB: Starting recovery for XA transactions... |
2021-12-26 19:39:04 0 [Note] InnoDB: Transaction 126555 in prepared state after recovery |
2021-12-26 19:39:04 0 [Note] InnoDB: Transaction contains changes to 1 rows |
2021-12-26 19:39:04 0 [Note] InnoDB: 1 transactions in prepared state after recovery |
2021-12-26 19:39:04 0 [Note] Found 1 prepared transaction(s) in InnoDB |
2021-12-26 19:39:04 0 [Note] Crash table recovery finished. |
2021-12-26 19:39:04 0 [Warning] Found prepared transaction with xid 118560 |
2021-12-26 19:39:04 0 [Warning] Could not complete 1 number of transactions. |
2021-12-26 19:39:04 0 [Note] InnoDB: Starting recovery for XA transactions... |
2021-12-26 19:39:04 0 [Note] InnoDB: Transaction 126555 in prepared state after recovery |
2021-12-26 19:39:04 0 [Note] InnoDB: Transaction contains changes to 1 rows |
2021-12-26 19:39:04 0 [Note] InnoDB: 1 transactions in prepared state after recovery |
2021-12-26 19:39:04 0 [Note] Found 1 prepared transaction(s) in InnoDB |
2021-12-26 19:39:04 0 [ERROR] Found 1 prepared transactions! It means that server was not shut down properly last time and critical recovery information (last binlog or tc.log file) was manually deleted after a crash. You have to start server with --tc-heuristic-recover switch to commit or rollback pending transactions. |
2021-12-26 19:39:04 0 [ERROR] Aborting |
And the binlog files are like this:
[19:23:09][root]$ ls -alt |
total 13336 |
drwxr-xr-x 2 rdsdb rdsdb 4096 Dec 26 19:18 . |
-rw-rw-rw- 1 rdsdb rdsdb 343 Dec 26 19:18 mysql-bin-changelog.index |
-rw-rw---- 1 rdsdb rdsdb 67146 Dec 20 20:25 mysql-bin-changelog.000021 |
-rw-rw---- 1 rdsdb rdsdb 1707199 Dec 20 20:25 mysql-bin-changelog.000020 |
-rw-rw---- 1 rdsdb rdsdb 4792114 Dec 20 20:25 mysql-bin-changelog.000019 |
-rw-rw-r-- 1 rdsdb rdsdb 7051915 Dec 20 20:23 mysql-bin-changelog.000018 |
-rw-rw-r-- 1 rdsdb rdsdb 717 Dec 20 20:20 mysql-bin-changelog.000017 |
-rw-rw-r-- 1 rdsdb rdsdb 717 Dec 20 20:15 mysql-bin-changelog.000016 |
-rw-rw-r-- 1 rdsdb rdsdb 717 Dec 20 20:10 mysql-bin-changelog.000015 |
drwxr-xr-x 10 rdsdb rdsdb 4096 Dec 20 19:04 .. |
[19:23:11][root]$ |
[19:23:12][root]$ /rdsdbbin/mysql/bin/mysqlbinlog mysql-bin-changelog.000021 | grep checkpoint |
#211220 20:25:29 server id 1604102129 end_log_pos 352 CRC32 0x1c7cf816 Binlog checkpoint mysql-bin-changelog.000020 |
#211220 20:25:29 server id 1604102129 end_log_pos 627 CRC32 0x6509ef3b Binlog checkpoint mysql-bin-changelog.000021 |
[19:23:15][root]$ |
I'll be OOO until January 3rd, later I can do more tests and private builds if needed. Happy holidays and happy new year!
So it looks very much that the block of static void xarecover_do_commit_or_rollback()
{
|
rc= xarecover_decide_to_commit(member, ptr_commit_max) ?
|
hton->commit_by_xid(hton, &x) : hton->rollback_by_xid(hton, &x);
|
|
|
/* |
It's fine to have non-zero rc which would be from transaction
|
non-participant hton:s.
|
*/
|
DBUG_ASSERT(rc || member->in_engine_prepare > 0); |
|
computes rc >0. It could be XAER_NOTA = -4 or XAER_RMFAIL = -7.
Also consider the latter code is returned "normally" at times:
innobase_commit_by_xid(
|
/*===================*/
|
handlerton* hton,
|
XID* xid) /*!< in: X/Open XA transaction identification */ |
{
|
...
|
if (high_level_read_only) { |
return(XAER_RMFAIL); |
}
|
Enjoy your yearend, bloody well deserved 
& Happy New one ahead!
Andrei
I tried several ways to reproduce the issue but seems it's not easy to reproduce with normal pkill commands. Probably requires a file system snapshot or something, I'm still testing.
However I created a tar file of the crashed data directory and I just sent to you in a separate email. Would you be able to help debugging with those data files?
Thanks,
Hugo
After debugging line by line using GDB, root cause was found.
When doing the recovery, the server generates a list of XIDs to be committed or rolled-back.
Then the callback function of innodb will be called to handle the commit or rollback.
In this case, the innodb transaction needs to be rolled-back. However it failed when searching the XID from trx_sys in function trx_get_trx_by_xid_callback . To be specific, failed when comparing xid data in line-2014. Code
Failed reason is server id doesn't match. Server ID is saved in xid_t.data[]. The format of xid_t.data[] is:
MySQLXid<Server ID 8 bytes><XID 8 bytes> |
Paste the GDB output here:
arg->xid is the input of the function to be searched from the hash.
element->trx->xid is the one exists in the hash which expected to match
Apparently gtrid_length, bqual_length, and transaction id all match but server ids are different.
- ((ulonglong)(arg->xid->data+8)) = 1049637262 which is the server id of the restored database.
- ((ulonglong)(element->trx->xid->data+8)) = 1604102129 which is the server id of the original database that the snapshot taken from.
(gdb) p element->id
|
$2 = 126555 |
(gdb) p element->trx
|
$3 = (trx_t *) 0x7fffebcca288 |
(gdb) p element->trx->is_recovered
|
$4 = true |
(gdb) p element->trx->state
|
$5 = {m = {static _S_min_alignment = 4, static _S_alignment = 4, _M_i = TRX_STATE_PREPARED}} |
(gdb) p arg->xid
|
$6 = (const XID *) 0x7fffffffad90 |
(gdb) p element->trx->xid
|
$7 = (XID *) 0x7ffff68b9258 |
(gdb) p arg->xid->gtrid_length
|
$8 = 24 |
(gdb) p element->trx->xid->gtrid_length
|
$9 = 24 |
(gdb) p arg->xid->bqual_length
|
$10 = 0 |
(gdb) p element->trx->xid->bqual_length
|
$11 = 0 |
(gdb) p arg->xid->data
|
$12 = "MySQLXid\216\061\220>\000\000\000\000 \317\001", '\000' <repeats 22 times>, "\002\000\000\000\000\000\000\000\277;W\031\000\000\000`\256\377\377\377\177\000\000J\006\001\000\000\000\000\000\023", '\000' <repeats 15 times>, "й\377\377\377\177\000\000\266\071\000\000\000\000\000\000Ю\377\377\377\177\000\000\337\361wVUU\000\000\230\000\000\000\000\000\000" |
(gdb) p element->trx->xid->data
|
$13 = "MySQLXid\361\247\234_\000\000\000\000 \317\001", '\000' <repeats 108 times> |
(gdb) p *((ulonglong*)(arg->xid->data+16)) |
$14 = 118560 |
(gdb) p *((ulonglong*)(element->trx->xid->data+16)) |
$15 = 118560 |
(gdb) p *((ulonglong*)(arg->xid->data+8)) |
$16 = 1049637262 |
(gdb) p *((ulonglong*)(element->trx->xid->data+8)) |
$17 = 1604102129 |
|
Why the server id doesn't match ?
After this commit, the input xid creation becomes using the set function:
void set(ulonglong xid) |
{
|
my_xid tmp;
|
formatID= 1; |
set(MYSQL_XID_PREFIX_LEN, 0, MYSQL_XID_PREFIX); |
memcpy(data+MYSQL_XID_PREFIX_LEN, &server_id, sizeof(server_id));
|
tmp= xid;
|
memcpy(data+MYSQL_XID_OFFSET, &tmp, sizeof(tmp));
|
gtrid_length=MYSQL_XID_GTRID_LEN;
|
}
|
This set function doesn't copy the original server id, instead it copy the new server_id of the new database.
memcpy(data+MYSQL_XID_PREFIX_LEN, &server_id, sizeof(server_id))
|
Therefore the xid can't be find in trx_sys because the server id doesn't match any more.
Why there's no such issue in 10.5?
In 10.5, the input xid was created directly from original xid : code xid= *trx->xid;
XID& xid= arg->xid_list[arg->count];
|
if (arg->count++ < arg->len) |
{
|
trx->state= TRX_STATE_PREPARED_RECOVERED;
|
...
|
xid= *trx->xid;
|
}
|
So the server match keeps the same with the old one.
To make it short, the issue could happen when restoring a new instance from original crashed data directory USING NEW SERVER ID. The issue doesn't exist in previous major versions before 10.6.
How to fix?
Seems a possible fix is to update the logic of creating the input XID object:
- Add original server id in struct xid_recovery_member.
- Then update the xarecover_do_commit_or_rollback function to set the original server id properly.
Proposed fix:
I've tested following changes fixed the issue. with the fix db recovered successfully.
While I'm still running the full mtr tests against the clean upstream commits to make sure it doesn't break anything, I paste it here as a raw diff file
If there's no breaking mtr tests and no obvious issues I will create PR to fix it.
Also may worth to add a MTR test, but it's late today I'll work on it tomorrow.
Elkin, could you please help review if the root cause and fix makes sense?
[wenhug@dev-dsk-wenhug-2b-f388f48d RDSMariaDB]$ cat brazil-build/patches/mariadb-1335.diff |
diff --git a/sql/handler.cc b/sql/handler.cc
|
index e4543bde5de..5cad422f45f 100644 |
--- a/sql/handler.cc
|
+++ b/sql/handler.cc
|
@@ -2414,7 +2414,7 @@ struct xarecover_st |
*/
|
static xid_recovery_member* |
xid_member_insert(HASH *hash_arg, my_xid xid_arg, MEM_ROOT *ptr_mem_root,
|
- XID *full_xid_arg)
|
+ XID *full_xid_arg, ulong server_id_arg)
|
{
|
xid_recovery_member *member= (xid_recovery_member *)
|
alloc_root(ptr_mem_root, sizeof(xid_recovery_member));
|
@@ -2428,7 +2428,7 @@ xid_member_insert(HASH *hash_arg, my_xid xid_arg, MEM_ROOT *ptr_mem_root, |
|
if (full_xid_arg) |
*xid_full= *full_xid_arg;
|
- *member= xid_recovery_member(xid_arg, 1, false, xid_full); |
+ *member= xid_recovery_member(xid_arg, 1, false, xid_full, server_id_arg); |
|
return |
my_hash_insert(hash_arg, (uchar*) member) ? NULL : member;
|
@@ -2443,14 +2443,15 @@ xid_member_insert(HASH *hash_arg, my_xid xid_arg, MEM_ROOT *ptr_mem_root, |
*/
|
static bool xid_member_replace(HASH *hash_arg, my_xid xid_arg, |
MEM_ROOT *ptr_mem_root,
|
- XID *full_xid_arg)
|
+ XID *full_xid_arg,
|
+ ulong server_id_arg)
|
{
|
xid_recovery_member* member;
|
if ((member= (xid_recovery_member *) |
my_hash_search(hash_arg, (uchar *)& xid_arg, sizeof(xid_arg))))
|
member->in_engine_prepare++;
|
else |
- member= xid_member_insert(hash_arg, xid_arg, ptr_mem_root, full_xid_arg);
|
+ member= xid_member_insert(hash_arg, xid_arg, ptr_mem_root, full_xid_arg, server_id_arg);
|
|
return member == NULL; |
}
|
@@ -2502,7 +2503,7 @@ static void xarecover_do_commit_or_rollback(handlerton *hton, |
Binlog_offset *ptr_commit_max= arg->binlog_coord;
|
|
if (!member->full_xid) |
- x.set(member->xid);
|
+ x.set(member->xid, member->server_id);
|
else |
x= *member->full_xid;
|
|
@@ -2660,7 +2661,8 @@ static my_bool xarecover_handlerton(THD *unused, plugin_ref plugin, |
{
|
// remember "full" xid too when it's not in mysql format |
if (xid_member_replace(info->commit_list, x, info->mem_root, |
- is_server_xid? NULL : &info->list[i]))
|
+ is_server_xid? NULL : &info->list[i],
|
+ is_server_xid? info->list[i].get_trx_server_id(): server_id))
|
{
|
info->error= true; |
sql_print_error("Error in memory allocation at xarecover_handlerton"); |
diff --git a/sql/handler.h b/sql/handler.h
|
index 0cdaf1dd698..73014093705 100644 |
--- a/sql/handler.h
|
+++ b/sql/handler.h
|
@@ -895,6 +895,18 @@ struct xid_t { |
memcpy(data+MYSQL_XID_OFFSET, &tmp, sizeof(tmp));
|
gtrid_length=MYSQL_XID_GTRID_LEN;
|
}
|
+
|
+ void set(ulonglong xid, ulong server_id) |
+ {
|
+ my_xid tmp;
|
+ formatID= 1; |
+ set(MYSQL_XID_PREFIX_LEN, 0, MYSQL_XID_PREFIX); |
+ memcpy(data+MYSQL_XID_PREFIX_LEN, &server_id, sizeof(server_id));
|
+ tmp= xid;
|
+ memcpy(data+MYSQL_XID_OFFSET, &tmp, sizeof(tmp));
|
+ gtrid_length=MYSQL_XID_GTRID_LEN;
|
+ }
|
+
|
void set(long g, long b, const char *d) |
{
|
formatID= 1; |
@@ -916,6 +928,12 @@ struct xid_t { |
!memcmp(data, MYSQL_XID_PREFIX, MYSQL_XID_PREFIX_LEN) ?
|
quick_get_my_xid() : 0; |
}
|
+ ulong get_trx_server_id()
|
+ {
|
+ ulong trx_server_id;
|
+ memcpy(&trx_server_id, data+MYSQL_XID_PREFIX_LEN, sizeof(server_id));
|
+ return trx_server_id; |
+ }
|
uint length()
|
{
|
return static_cast<uint>(sizeof(formatID)) + key_length(); |
@@ -957,11 +975,12 @@ struct xid_recovery_member |
bool decided_to_commit;
|
Binlog_offset binlog_coord; // semisync recovery binlog offset |
XID *full_xid; // needed by wsrep or past it recovery |
+ ulong server_id; // server id of original server |
|
xid_recovery_member(my_xid xid_arg, uint prepare_arg, bool decided_arg,
|
- XID *full_xid_arg)
|
+ XID *full_xid_arg, ulong server_id_arg)
|
: xid(xid_arg), in_engine_prepare(prepare_arg),
|
- decided_to_commit(decided_arg), full_xid(full_xid_arg) {};
|
+ decided_to_commit(decided_arg), full_xid(full_xid_arg) , server_id(server_id_arg) {};
|
};
|
|
/* for recover() handlerton call */ |
|
wenhug, thanks for digging it out!
Indeed, 6c39eaeb1 made recovery dependent on server_id.
Your diff makes sense, so feel free to go with a PR.
To a lower level, having only browsed it, I am not sure if a new full-blown SET method is necessary (while we only need server_id to overwrite)...
Cheers,
Andrei.
Elkin, I created a PR for the issue https://github.com/MariaDB/server/pull/1983 . Could you help review it ? Thanks a lot!
commit dafc5fb9c17637c6233e8a73c890f158d645a37c
Author: Hugo Wen <wenhug@amazon.com>
Date: Fri Feb 4 03:56:08 2022 +0000
is pushed into 10.6.