May 2023 - discuss - lists.mariadb.org

[Maria-discuss] Backup on the replication server getting affected
by ragul rangarajan 09 Jun '23

09 Jun '23

3 3

[Maria-discuss] jujst re-joined trial e-mail don't read
by antonio scotti 30 May '23

30 May '23

=============================================== PermaCultura-es http://www.permacultura-es.org

1 0

[Maria-discuss] trial e-mail don't read
by antonio scotti 30 May '23

30 May '23

=============================================== PermaCultura-es http://www.permacultura-es.org

1 0

[Maria-discuss] Innodb_buffer_pool_read_requests is zero
by mariadb＠Biblestuph.com 19 May '23

19 May '23

Just upgraded from 10.6.12 to 10.6.13 on a Debian Bullseye system, and seeing this oddity. Innodb_buffer_pool_read_requests is zero, and seems to remain at zero no matter what I do (I even tried selecting all the data from a small table several times in a row with no change). It's been running a couple hours now; I've restarted it a couple times. I have 10.6.13 on a few other platforms as well (including Buster), and I'm not seeing the same issue on them. What would cause this? I have no other Bullseye systems running 10.6.13 yet, and the same system has been running 10.6.12 for over a week now without showing this issue. The system _seems_ to be running fine, it's a replica system that is being fed actively from the master (data comes in to it very regularly). MariaDB [(none)]> SHOW GLOBAL STATUS LIKE 'innodb_buffer_pool_%'; +-----------------------------------------+--------------------------------------------------+ | Variable_name | Value | +-----------------------------------------+--------------------------------------------------+ | Innodb_buffer_pool_dump_status | | | Innodb_buffer_pool_load_status | Buffer pool(s) load completed at 230511 23:21:26 | | Innodb_buffer_pool_resize_status | | | Innodb_buffer_pool_load_incomplete | OFF | | Innodb_buffer_pool_pages_data | 609103 | | Innodb_buffer_pool_bytes_data | 9979543552 | | Innodb_buffer_pool_pages_dirty | 19465 | | Innodb_buffer_pool_bytes_dirty | 318914560 | | Innodb_buffer_pool_pages_flushed | 0 | | Innodb_buffer_pool_pages_free | 299441 | | Innodb_buffer_pool_pages_made_not_young | 0 | | Innodb_buffer_pool_pages_made_young | 107 | | Innodb_buffer_pool_pages_misc | 0 | | Innodb_buffer_pool_pages_old | 224824 | | Innodb_buffer_pool_pages_total | 908544 | | Innodb_buffer_pool_pages_lru_flushed | 0 | | Innodb_buffer_pool_pages_lru_freed | 0 | | Innodb_buffer_pool_pages_split | 726 | | Innodb_buffer_pool_read_ahead_rnd | 0 | | Innodb_buffer_pool_read_ahead | 3 | | Innodb_buffer_pool_read_ahead_evicted | 0 | | Innodb_buffer_pool_read_requests | 0 | | Innodb_buffer_pool_reads | 604346 | | Innodb_buffer_pool_wait_free | 0 | | Innodb_buffer_pool_write_requests | 3639341 | +-----------------------------------------+--------------------------------------------------+ 25 rows in set (0.000 sec) MariaDB [(none)]> SHOW VARIABLES LIKE 'innodb_buffer_pool%'; +-------------------------------------+----------------+ | Variable_name | Value | +-------------------------------------+----------------+ | innodb_buffer_pool_chunk_size | 134217728 | | innodb_buffer_pool_dump_at_shutdown | ON | | innodb_buffer_pool_dump_now | OFF | | innodb_buffer_pool_dump_pct | 80 | | innodb_buffer_pool_filename | ib_buffer_pool | | innodb_buffer_pool_load_abort | OFF | | innodb_buffer_pool_load_at_startup | ON | | innodb_buffer_pool_load_now | OFF | | innodb_buffer_pool_size | 15032385536 | +-------------------------------------+----------------+ 9 rows in set (0.000 sec)

2 5

[Maria-discuss] Question regarding mariadb-dump/mysqldump and rocksdb
by Jonas Krauss 16 May '23

16 May '23

Hi MariaDB, in Myrocks / Facebook MySQL there was a useful addition to mysqldump which makes queries issued by the dumper not add to the block cache of the rocksdb instance. It was added quite some time ago: https://github.com/facebook/mysql-5.6/commit/2223479fea78a246884c5a09ef4d02… The corresponding server variable used is *rocksdb_skip_fill_cache* Is this active in MariaDB's version of mysqldump as well? The server variable exists in MariaDB's MyRocks (checked for 10.6.12) but I have not found a reference in MariaDB's source code when searching for rocksdb_skip_fill_cache. If it is not active it would be great if it could be added in a future version. Best Jonas

2 2

[Maria-discuss] Undo logs aren't truncating
by mariadb＠Biblestuph.com 15 May '23

15 May '23

On a recently updated MariaDB 10.6.13 server (Debian Bullseye) I noticed that my ibdata1 file was continuously growing; it had gotten a little over 3G after running a couple days. I would rather keep this growth in check so used it as an excuse to setup undo tablespaces that could be truncated (I had wanted to try this for some time anyway). So: Dump the whole DB, delete all the InnoDB tables, stop the server, delete the innodb system tables, reset the appropriate undo variables (with innodb_max_undo_log_size set to 750M, innodb_undo_tablespaces set to 4, and innodb_undo_log_truncate set to ON), restart the server and reload the data. I now have FOUR undo files that appear to be continuously growing; as of this writing they are over 1G and still increasing. This is an active read-only replica (processing updates very frequently; at times continuously but not constantly as such). How do I get it to actually reclaim the space? Or am I misunderstanding how this is supposed to work? Below are my current relevant settings (or at least what I think are relevant). Normally I run with innodb_purge_rseg_truncate_frequency = 128; but for testing I have it now set to 1 (but with no noticeable difference). Things I have tried (no luck on any of them): Set innodb_purge_rseg_truncate_frequency = 1. Set innodb_max_purge_lag_wait=0 (it returned nearly immediately). Stop both slave threads and wait for several minutes. Set innodb_max_purge_lag and innodb_max_purge_lag_delay (see below). MariaDB [(none)]> show variables like '%trunc%'; +--------------------------------------+-------+ | Variable_name | Value | +--------------------------------------+-------+ | innodb_purge_rseg_truncate_frequency | 1 | | innodb_undo_log_truncate | ON | +--------------------------------------+-------+ 2 rows in set (0.000 sec) MariaDB [(none)]> show variables like '%purge%'; +--------------------------------------+------------+ | Variable_name | Value | +--------------------------------------+------------+ | aria_log_purge_type | immediate | | innodb_max_purge_lag | 1000 | | innodb_max_purge_lag_delay | 5000 | | innodb_max_purge_lag_wait | 4294967295 | | innodb_purge_batch_size | 300 | | innodb_purge_rseg_truncate_frequency | 1 | | innodb_purge_threads | 4 | | relay_log_purge | ON | +--------------------------------------+------------+ 8 rows in set (0.000 sec) MariaDB [(none)]> show variables like '%undo%'; +--------------------------+-----------+ | Variable_name | Value | +--------------------------+-----------+ | innodb_max_undo_log_size | 786432000 | | innodb_undo_directory | ./ | | innodb_undo_log_truncate | ON | | innodb_undo_tablespaces | 4 | +--------------------------+-----------+ 4 rows in set (0.000 sec)

2 1

[Maria-discuss] mariadb crash und recovery
by Radisson97＠gmx.de 14 May '23

14 May '23

Hello list, i had a crash at weekend with no apparend cause. The aria engine did a recovery without furture problems. Maybe that is a hint that somebody can make use of. Hints welcome. CU from log: 230416 7:20:17 [ERROR] mysqld got signal 11 ; This could be because you hit a bug. It is also possible that this binary or one of the libraries it was linked against is corrupt, improperly built, or misconfigured. This error can also be caused by malfunctioning hardware. To report this bug, see https://mariadb.com/kb/en/reporting-bugs We will try our best to scrape up some info that will hopefully help diagnose the problem, but since we have already crashed, something is definitely wrong and this may fail. Server version: 10.5.15-MariaDB-0+deb11u1-log key_buffer_size=134217728 read_buffer_size=131072 max_used_connections=86 max_threads=252 thread_count=38 It is possible that mysqld could use up to key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 685818 K bytes of memory Hope that's ok; if not, decrease some variables in the equation. Thread pointer: 0x7f01e4000c58 Attempting backtrace. You can use the following information to find out where mysqld died. If you see no messages after this, something went terribly wrong... stack_bottom = 0x7f02b6894d38 thread_stack 0x49000 ??:0(my_print_stacktrace)[0x5559d92bb54e] ??:0(handle_fatal_signal)[0x5559d8dbaf65] sigaction.c:0(__restore_rt)[0x7f043591a140] ??:0(_ma_state_info_write)[0x5559d8fafc73] ??:0(_ma_state_info_write)[0x5559d8fb1737] ??:0(_ma_state_info_write)[0x5559d8fb946c] ??:0(handler::ha_write_row(unsigned char const*))[0x5559d8dc7aa8] ??:0(MyCTX_nopad::finish(unsigned char*, unsigned int*))[0x5559d8ff40da] ??:0(handler::ha_write_row(unsigned char const*))[0x5559d8dc7aa8] ??:0(write_record(THD*, TABLE*, st_copy_info*, select_result*))[0x5559d8b8787d] ??:0(mysql_insert(THD*, TABLE_LIST*, List<Item>&, List<List<Item> >&, List<Item>&, List<Item>&, enum_duplicates, bool, select _result*))[0x5559d8b8f8a4] ??:0(mysql_execute_command(THD*))[0x5559d8bc2716] ??:0(mysql_parse(THD*, char*, unsigned int, Parser_state*, bool, bool))[0x5559d8bc65db] ??:0(dispatch_command(enum_server_command, THD*, char*, unsigned int, bool, bool))[0x5559d8bc8a5d] ??:0(do_command(THD*))[0x5559d8bca2de] ??:0(do_handle_one_connection(CONNECT*, bool))[0x5559d8cbbfb2] ??:0(handle_one_connection)[0x5559d8cbc22d] ??:0(MyCTX_nopad::finish(unsigned char*, unsigned int*))[0x5559d8ff811b] nptl/pthread_create.c:478(start_thread)[0x7f043590eea7] x86_64/clone.S:97(__GI___clone)[0x7f0435525def] Trying to get some variables. Some pointers may be invalid and cause the dump to abort. Query (0x7f01e4010440): INSERT INTO messwerte VALUES ( 83110995 , 1200 , 909 , 331 , '2023-04-16 05:20:00' , '2023-04-16 05: 20:00' , 0.0703000000000000013766765505351941101253032684326171875 , 0.001000000000000000020816681711721685132943093776702880 859375 , 0 , 0 , 0 , 0 , 60 , 0 ) Connection ID (thread ID): 5463672 Status: NOT_KILLED Optimizer switch: index_merge=on,index_merge_union=on,index_merge_sort_union=on,index_merge_intersection=on,index_merge_sort_ intersection=off,engine_condition_pushdown=off,index_condition_pushdown=on,derived_merge=on,derived_with_keys=on,firstmatch=o n,loosescan=on,materialization=on,in_to_exists=on,semijoin=on,partial_match_rowid_merge=on,partial_match_table_scan=on,subque ry_cache=on,mrr=off,mrr_cost_based=off,mrr_sort_keys=off,outer_join_with_cache=on,semijoin_with_cache=on,join_cache_increment al=on,join_cache_hashed=on,join_cache_bka=on,optimize_join_buffer_size=on,table_elimination=on,extended_keys=on,exists_to_in= on,orderby_uses_equalities=on,condition_pushdown_for_derived=on,split_materialized=on,condition_pushdown_for_subquery=on,rowi d_filter=on,condition_pushdown_from_having=on,not_null_range_scan=off The manual page at https://mariadb.com/kb/en/how-to-produce-a-full-stack-trace-for-mysqld/ contains information that should help you find out what is causing the crash. Writing a core file... Working directory at /home/kpc/kpc.dbs Resource Limits: Limit Soft Limit Hard Limit Units Max cpu time unlimited unlimited seconds Max file size unlimited unlimited bytes Max data size unlimited unlimited bytes Max stack size 8388608 unlimited bytes Max core file size 0 unlimited bytes Max resident set unlimited unlimited bytes Max processes 127976 127976 processes Max open files 32313 32313 files Max locked memory 65536 65536 bytes Max address space unlimited unlimited bytes Max file locks unlimited unlimited locks Max pending signals 127976 127976 signals Max msgqueue size 819200 819200 bytes Max nice priority 0 0 Max realtime priority 0 0 Max realtime timeout unlimited unlimited us Core pattern: core 230416 07:20:19 mysqld_safe Number of processes running now: 0 230416 07:20:19 mysqld_safe mysqld restarted 2023-04-16 7:20:19 0 [Note] /usr/sbin/mariadbd (mysqld 10.5.15-MariaDB-0+deb11u1-log) starting as process 415 ... 2023-04-16 7:20:20 0 [Note] mariadbd: Aria engine: starting recovery recovered pages: 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% (0.0 seconds); tables to flush: 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 (0.1 seconds); 2023-04-16 7:20:20 0 [Note] mariadbd: Aria engine: recovery done

4 3

[Maria-discuss] Checkpoint Age and Redo Log Flushing in 10.5+
by Gordan Bobic 14 May '23

14 May '23

Ever since 10.5, the redo log flushing seems to have become ... suboptimal. It looks like the database does nothing at all to flush the logs until it hits the hwm, and then it goes into overdrive and flushes everything to tablespaces fast, before it goes back to doing nothing until the log fills up. This causes a problem when the redo log is already full and a large burst of writes arrives because it overwhelms the redo log and causes a stall. Is there a bug fix for this or a configuration change available to restore the old behaviour where everything gets flushed continuously and switches up a gear when the hwm is reached? Here are the screenshots comparing 10.5+ sawtooth pattern vs 10.4 and older versions.

2 1

[Maria-discuss] howto c-connect bindings
by juerg oehler 06 May '23

06 May '23

hi I'm working with|mariadb|and|c-connect|prepared statements and run into crashes executing them once a while. i so addressed my problem to stakeoverflow <https://stackoverflow.com/questions/76170354/mariadb-howto-c-connect-bindin…> hoping to get more information for binding programs data structure to db interface. i did get an answer. but this answer does not go inline with your documentation <https://mariadb.com/kb/en/mariadb-connectorc-data-structures/#mysql_bind>. i so do have the option to follow your documentation or use the answer and run into crashes in either way. you might ask why i'm not just going with your documentation? your documentation is far to thin to get a reliable idea how bind structure is used and i do get proofed this by crashes. second i played with <my_bool * is_nulll Pointer which indicates if column is NULL (not used for parameters)> for parameters and behavior was not as documented. this is my motivation to bring my problem to this mailing list hopping someone can point me to a documentation i can rely on. i found a short and useful documentation in mysql with permissible in and out types <https://dev.mysql.com/doc/c-api/8.0/en/c-api-prepared-statement-type-codes.…> and statement data structures <https://dev.mysql.com/doc/c-api/8.0/en/c-api-prepared-statement-data-struct…>. can i use these informations to build 2 test applications one focusing on parameter (attachment) one focusing on result? currently i implemented MariaPrepUpd.c using mariadb documentation. i felt like flying blind and ended up in crashes - once a while. do you think i could use mysql documentation for improving bindings? or do you have a better option? (i studied your bulk example <https://mariadb.com/kb/en/bulk-insert-row-wise-binding/>. i even was using this example as base for first try an hit the wall by using STMT_INDICATOR_NTS. i then learned my problem is not just the example using STMT_ATTR_ARRAY_SIZE = 1.) my goal is to have a version of MariaPrepUpd.c i can use for any application for iteration 1 to n. thank you for short answer. have a great day. juerg(=george) oehler

1 0