November 2013 - developers - lists.mariadb.org

Re: [Maria-developers] Rev 3908: MDEV-5277 - Ensure that all MySQL 5.6 options are supported
by Sergei Golubchik 25 Nov '13

25 Nov '13

Hi, Sergey! On Nov 20, Sergey Vojtovich wrote: > revno: 3908 > revision-id: svoj(a)mariadb.org-20131120113753-vuxj3v4jt9cvccdy > parent: svoj(a)mariadb.org-20131114140000-5eq8gopqu5giwvqq > committer: Sergey Vojtovich <svoj(a)mariadb.org> > branch nick: 10.0-mdev5277 > timestamp: Wed 2013-11-20 15:37:53 +0400 > message: > MDEV-5277 - Ensure that all MySQL 5.6 options are supported by the > MariaDB 10.0 server > === modified file 'mysql-test/r/mysqld--help.result' > --- a/mysql-test/r/mysqld--help.result 2013-11-03 20:26:44 +0000 > +++ b/mysql-test/r/mysqld--help.result 2013-11-20 11:37:53 +0000 > @@ -4,6 +4,9 @@ > --defaults-file=# Only read default options from the given file #. > --defaults-extra-file=# Read this file after the global files are read. > > + --abort-slave-event-count[=name] > + MySQL 5.6 compatibility option. Use > + '--debug-abort-slave-event-count' instead. > --allow-suspicious-udfs Do we really want all these do-nothing compatibility options to show up in --help? May be not? The goal of this MDEV was not to fail with an error when starting on mysqld's my.cnf file, no one needs --help for that. > Allows use of UDFs consisting of only one symbol xxx() > without corresponding xxx_init() or xxx_deinit(). That > @@ -136,10 +151,16 @@ > --deadlock-timeout-short=# > Short timeout for the two-step deadlock detection (in > microseconds) > + --default-authentication-plugin[=name] > + MySQL 5.6 compatibility option. To be implemented in > + later 10.x versions. Remove "10.x" please, just "in later versions". We don't want to promise that all these options will be implemented in 10.x series. > === modified file 'storage/innobase/handler/ha_innodb.cc' > --- a/storage/innobase/handler/ha_innodb.cc 2013-11-13 21:58:19 +0000 > +++ b/storage/innobase/handler/ha_innodb.cc 2013-11-20 11:37:53 +0000 > @@ -15750,6 +15750,12 @@ static MYSQL_SYSVAR_BOOL(purge_stop_now, > PLUGIN_VAR_OPCMDARG, > "Set purge state to STOP", > NULL, purge_stop_now_set, FALSE); > + > +static my_bool innodb_log_checkpoint_now; > +static MYSQL_SYSVAR_BOOL(log_checkpoint_now, innodb_log_checkpoint_now, > + PLUGIN_VAR_OPCMDARG | PLUGIN_VAR_NOSYSVAR, > + "MySQL 5.6 compatibility option. To be implemented in later 10.x versions.", > + NULL, NULL, FALSE); > #endif /* UNIV_DEBUG */ Why is that (and everything else in ha_innodb.cc) necessary? We'll merge with the latest InnoDB, we will have all its variables. Regards, Sergei

2 3

[Maria-developers] MariaDB-10.0-beta sysbench results
by Axel Schwenke 23 Nov '13

23 Nov '13

Hi, I have run another sysbench with in-memory workload to check where we are with respect to mutex contention. The benchmark is sysbench OLTP (using the bzr trunk of sysbench with LUA support). 10 mio rows in a single table (about 2.5G .ibd file). InnoDB buffer pool set to 32G and warmed up. I have run the test on two different machines. First on our old trusty facebook2 machine with 8 cores/16 threads in two NUMA zones. Second on the fat lizard2 machine with 32 cores/64 threads in 4 NUMA zones. As expected, the heavy contention of the LOCK_open mutex is more visible on the bigger machine (I have not done any detailed analysis in this benchmark, but before while doing benchmarks for MDEV-5081) I notice a visible performance gain in MariaDB-10.0 over MariaDB-5.5 at 32 threads. This is probably due to the work on the InnoDB kernel mutex (impossible to prove since 5.5 doesn't have mutex instrumentation). At higher concurrency however the LOCK_open mutex kicks in and performance is clearly dominated by that mutex alone (it flattens out for all 5.5 and 10.0 versions of MariaDB) The good news is, that 10.0.4, 10.0.5 and 10.0.6 are quite close. There is a small regression from 10.0.4 to 10.0.5, but IMHO this isn't worth investigation while the LOCK_open issue is not fixed. Attached: diagrams. OpenOffice spreadsheets. XL

6 7

[Maria-developers] get is now a reserved word in mariadb 10 ?
by Jocelyn Fournier 21 Nov '13

21 Nov '13

Hi, I have a query with a column named "get". i.e. SELECT get FROM my_table... It works perfectly in 5.5, but in 10.0.6 it fails : ERROR 1064 (42000): You have an error in your SQL syntax; check the manual that corresponds to your MariaDB server version for the right syntax to use near 'get FROM my_table' at line 1 I assume "get" is a new reserved word, probably inherited from MySQL 5.6 memcached support, although memcached is not supported in MariaDB ? (it's mentionned here : http://dev.mysql.com/doc/mysqld-version-reference/en/mysqld-version-referen… ) Is there a specific KB entry for MariaDB 10.0.6 reserved words ? (I only see an entry for 5.5 here : https://mariadb.com/kb/en/reserved-words/ ) Thanks and regards, Jocelyn Fournier

4 4

[Maria-developers] Prep for MariaDB 5.5.34
by Daniel Bartholomew 20 Nov '13

20 Nov '13

I'm currently in the process of preparing for the MariaDB 5.5.34 release. Draft changelog and release notes are here: - https://mariadb.com/kb/en/mariadb-5534-release-notes/ - https://mariadb.com/kb/en/mariadb-5534-changelog/ As always, fixes, additions, and such are welcome. Buildbot is still working on building the packages and when it's finished I'll get them prepped, signed, and uploaded to the mirrors with the plan to release tomorrow morning US Eastern Time. Thanks. -- Daniel Bartholomew MariaDB

1 0

[Maria-developers] MariaDB 10.0.6 using more file descriptors than 5.5 ?
by Jocelyn Fournier 20 Nov '13

20 Nov '13

Hi, I've just migrated MariaDB (+tokudb) from 5.5.33a to 10.0.6, and I've encountered a '[ERROR] Error in accept: Too many open files' error during a backup. As this was not occuring with MariaDB 5.5, is this expected to have 10.0.6 using much more file descriptors ? (open_files_limit was set to about 8100 with 5.5, I raised it to 65535 now to fix this issue). Thanks, Jocelyn Fournier

3 3

[Maria-developers] MDEV-5277 - Ensure that all MySQL 5.6 options are supported by the MariaDB 10.0 server
by Sergey Vojtovich 20 Nov '13

20 Nov '13

Hi! I'm gathering 5.6 options that are not present in 10.0 and the way we want to handle them. Could you check options listed below and if you're aware of their destiny indicate if options: * not used or needed in MariaDB * to be implemented in later 10.x versions * tuned via different 10.x specific options The following options exist in 5.6 but not in 10.0: * default-tmp-storage-engine * log-raw * default_authentication_plugin * ignore-db-dir * binlog-max-flush-queue-time * binlog-row-image * explicit-defaults-for-timestamp * master-info-repository * relay-log-info-repository * binlog-rows-query-log-events * binlog-order-commits * log-throttle-queries-not-using-indexes * end-markers-in-json * optimizer-trace * optimizer-trace-features * optimizer-trace-offset * optimizer-trace-limit * optimizer-trace-max-mem-size * eq-range-index-dive-limit * server-uuid * server-id-bits * slave-rows-search-algorithms * table-open-cache-instances * slave-allow-batching * slave-checkpoint-period * slave-checkpoint-group * slave-parallel-workers * slave-pending-jobs-size-max * enforce-gtid-consistency * gtid-mode * disconnect-on-expired-password * sha256-password-private-key-path (sha256_password plugin) * sha256-password-public-key-path (sha256_password plugin) The following options exist in 5.5 and 5.6 but not in 10.0: * abort-slave-event-count => debug-abort-slave-event-count * disconnect-slave-event-count => debug-disconnect-slave-event-count * exit-info => debug-exit-info * max-binlog-dump-events => debug-max-binlog-dump-events * sporadic-binlog-dump-fail => debug-sporadic-binlog-dump-fail * new * skip-new * skip-stack-trace The following options were added after 5.6.10: * innodb-buffer-pool-evict // UNIV_DEBUG * innodb-ft-total-cache-size * innodb-ft-result-cache-limit * innodb-log-compressed-pages * innodb-log-checkpoint-now // UNIV_DEBUG * rpl-stop-slave-timeout * validate-user-plugins The following NO_CMD_LINE options exist in 5.6 but not in 10.0 (won't add them): * gtid-executed * gtid-next * gtid-next-list * gtid-owned * gtid-purged * log-bin-basename * relay-log-basename * transaction-allow-batching Thanks, Sergey

4 12

Re: [Maria-developers] Rev 3974: MDEV-5182 - build of 10.0.4/r3863 fails @ 'cmake'
by Sergei Golubchik 20 Nov '13

20 Nov '13

Hi, Sergey! On Nov 18, Sergey Vojtovich wrote: > ------------------------------------------------------------ > revno: 3974 > revision-id: svoj(a)mariadb.org-20131118114801-1kny2cvx8kpjsw7t > parent: jplindst(a)mariadb.org-20131115132442-5d4hqg5idjta2dni > committer: Sergey Vojtovich <svoj(a)mariadb.org> > branch nick: 5.5-mdev5182 > timestamp: Mon 2013-11-18 15:48:01 +0400 > message: > MDEV-5182 - build of 10.0.4/r3863 fails @ 'cmake' with > -DINSTALL_SYSCONFDIR/-DDEFAULT_SYSCONFDIR specified > > The reason for a bug is in support-files/CMakeLists.txt that tries to > install files in INSTALL_SYSCONF2DIR, without checking whether this > variable is actually defined. > > Normally INSTALL_SYSCONF2DIR is always defined if INSTALL_SYSCONFDIR is > defined, and support-files/CMakeLists.txt assumes that. But when one > specifies INSTALL_SYSCONFDIR manually on the command line, this > assumption might be broken. Looks ok to me. An alternative solution would be to define INSTALL_SYSCONF2DIR if INSTALL_SYSCONFDIR is defined. Regards, Sergei

2 1

[Maria-developers] Open Questions in the Knowledge Base
by Ian Gilfillan 19 Nov '13

19 Nov '13

There are quite a few unanswered questions in the Knowledge Base https://mariadb.com/kb/en/+questions/ If anyone knowledgeable in those areas can help answer some of these, even just a request for more info, that would be great :) thanks, ian

2 1

Re: [Maria-developers] Accessing rpl_global_gtid_binlog_state is not consistently protected
by Kristian Nielsen 19 Nov '13

19 Nov '13

Pavel Ivanov <pivanof(a)google.com> writes: > sql/log.cc has highly inconsistent pattern of accessing > rpl_global_gtid_binlog_state. In some places it's protected by mutex > rpl_global_gtid_binlog_state.LOCK_binlog_state, in some places it's > protected by global mutex LOCK_rpl_gtid_state, and in some places it's > not protected by any mutex at all. This is a bug that should be fixed, > right? Yes, agree, thanks for finding this. > I can file this in JIRA if you want, just wanted to let you know faster. I've filed it as MDEV-5306. > I've noticed this problem after one of our AddressSanitizer testing > bots reported that rpl_binlog_state::bump_seq_no_if_needed (called > from MYSQL_BIN_LOG::bump_seq_no_counter_if_needed) was using memory > from hash after it was already freed by rpl_binlog_state::reset > (called by MYSQL_BIN_LOG::reset_logs). Can you test again with the latest patch I've just pushed to 10.0-base? I looked briefly into making an mtr test case to catch this, but I did not see any easy way to do that. I went through the code to try and check all uses of rpl_binlog_state for proper locking, but it would be nice to get a confirmation that I did not miss anything. - Kristian.

3 2

Re: [Maria-developers] MariaDB Galera replication
by Pavel Ivanov 18 Nov '13

18 Nov '13

I'm starting a new thread as this is already doesn't have anything to do with the original topic. On Fri, Nov 15, 2013 at 10:46 AM, Alex Yurchenko <alexey.yurchenko(a)codership.com> wrote: >>> Please pardon this arrogant interruption of your discussion and shameless >>> self-promotion, but I just could not help noticing that Galera >>> replication >>> was designed specifically with these goals in mind. And it does seem to >>> achieve them better than semi-sync plugin. Have you considered Galera? >>> What >>> makes you prefer semi-sync over Galera, if I may ask? >> >> >> To be honest I never looked at how Galera works before. I've looked at >> it now and I don't see how it can fit with us. The major disadvantages >> I immediately see: >> 1. Synchronous replication. That means client must wait while >> transaction is applied on all nodes which is unacceptably big latency >> of each transaction. And what if there's a network blip and some node >> becomes inaccessible? All writes will just freeze? I see the statement >> that "failed nodes automatically excluded from the cluster", but to do >> that cluster must wait for some timeout in case it's indeed a network >> blip and node will "quickly" reconnect. And every client must wait for >> cluster to decide what happened with that one node. >> 2. Let's say node fell out of the cluster for 5 minutes and then >> reconnected. I guess it will be treated as "new node", it will >> generate state transfer and the node will start downloading the whole >> database? And while it's trying to download say 500GB of data files >> all other nodes (or maybe just donor?) won't be able to change those >> files locally and thus will blow up its memory consumption. That means >> they could quickly run out of memory and "new node" won't be able to >> finish its "initialization"... >> 3. It looks like there's strong asymmetry in starting cluster nodes -- >> the first one should be started with empty wsrep_cluster_address and >> all others should be started with the address of the first node. So I >> can't start all nodes uniformly and then issue some commands to >> connect them to each other. That's bad. >> 4. What's the transition path? How do I upgrade MySQL/MariaDB >> replicating using usual replication to Galera? It looks like there's >> no such path and the solution is stop the world using regular >> replication and restart it using Galera. Sorry I can't do that with >> our production systems. >> >> I believe these problems are severe enough for us, so that we can't >> work with Galera. > > > Pavel, you seem to be terribly mistaken on almost all accounts: > > 1. *Replication* (i.e. data buffer copying) is indeed synchronous. But > nobody said that commit is. What Galera does is very similar to semi-sync, > except that it does it technically better. I would not dare to suggest > Galera replication if I didn't believe it to be superior to semi-sync in > every respect. Well, apparently we have a different understanding of what the term "synchronous replication" means. This term is all over the Galera doc, but I didn't find the detailed description of how actually Galera replication work. So I assumed that my understanding of the term (which actually seem to be in line with wiki's definitions http://en.wikipedia.org/wiki/Replication_(computing) ) is what was implied there. So I hope you'll be able to describe in detail how Galera replication works. > As an example here's an independent comparison of Galera vs. > semi-sync performance: > http://linsenraum.de/erkules/2011/06/momentum-galera.html. This is a nice blog post written in German and posted in 2011. And while Google Translate gave me an idea what post was about it would be nice to see something more recent and with better description of what was the actual testing set up. > In fact, majority > of Galera users migrated from the regular *asynchronous* MySQL replication, > which I think is a testimony to Galera performance. I don't mean to troll, but this can also mean that everyone who migrated didn't care much about performance and Galera's performance was within sane boundaries... BTW, just found here https://mariadb.com/kb/en/mariadb-galera-cluster-known-limitations/ : "by design performance of the cluster cannot be higher than performance of the slowest node; however, even if you have only one node, its performance can be considerably lower comparing to running the same server in a standalone mode". That contradicts your words. > 2. Node reconnecting to cluster will normally receive only events that it > missed while being disconnected. This seem to contradict to the docs. Again from https://mariadb.com/kb/en/mariadb-galera-cluster-known-limitations/ : "After a temporary split, if the 'good' part of the cluster was still reachable and its state was modified, resynchronization occurs". > 3. You are partially right about it, but isn't it much different from > regular MySQL replication where you first need to set up master and then > connect slaves (even if you have physically launched the servers at the same > time). Operation of setting up master and then connecting slaves consists of mostly only executing CHANGE MASTER TO and then START SLAVE on all slaves after all MySQL instances (including master) were started with the same set of command line flags. This is fundamentally different from starting instances with different arguments, especially when these arguments should be different depending on whether the replica is starting first or there's already some other replica running. > Yet, Galera nodes can be started simultaneously and then joined > together by setting wsrep_cluster_address from mysql client connection. This > is not advertised method, because in that case state snapshot transfer can > be done only by mysqldump. If you set the address in advance, rsync or > xtrabackup can be used to provision the fresh node. This is of course better because I can start all instances with the same command line arguments. But transferring snapshot of a very big database using mysqldump, and causing the node that creates mysqldump to blow up memory consumption during the process, that is still a big problem. > 4. Every Galera node can perfectly work as either master or slave to native > MySQL replication. So migration path is quite clear. Nope, not clear yet. So I'll be able to upgrade all my MySQL instances to a Galera-supporting binary while they are replicating using standard MySQL replication. That's good. Now, how the Galera replication is turned on after that? What will happen if I just set wsrep_cluster_address address on all replicas? What will replicas do, and what will happen with the standard MySQL replication? > It is very sad that you happen to have such gross misconceptions about > Galera. If those were true, how would MariaDB Galera Cluster get paying > customers? Care to share some numbers? Like what's the rough amount of those paying customers? What size is the biggest installation -- number of clusters, replicas, highest QPS load? I'm not asking to share any confidential information, but the rough ballpark of the numbers would be helpful. > May be my reply will convince you to have a second look at it. > (In addition to the above Galera is fully multi-master, does parallel > applying and works great in WAN) I hope your explanation of how Galera replication work will help me understand how great it works over WAN and how you could make full multi-master work without fully synchronous replication in my understanding of that term. Pavel

3 9