#maria Log v0.1

logs Catalogue

This page loads from the db, refresh as and when. It displays 24 hrs
For older data http://marialog.archivist.info/previous.php, logs go back to 2009-01-18
Some user stats
Table logs
date_time
user
message
2016-04-28 03:10:53
jonrwads
Hello everybody, I have a bit of a problem I have a function that the first time it is run it takes 1 second and then every subsequent time takes around 60 seconds. I have disabled cache and set the size to 0 so cache should not be affecting it. Any ideas?
2016-04-28 03:17:52
dragonheart
jonrwads: show details.
2016-04-28 03:19:53
jonrwads
Ok so I have pasted the function and results here http://pastebin.com/hA1jBUNv
2016-04-28 03:20:43
jonrwads
The view the function uses is not pretty but is kind of out of scope due to the fact that the first time it seems to be performant enough but on the next call split seconds later it takes 60 times longer
2016-04-28 03:26:00
dragonheart
jonrwads: use EXPLAIN {query}
2016-04-28 03:26:39
dragonheart
also why are you doing DATE(Y) < DATE(Y)? why not X< Y
2016-04-28 03:27:10
dragonheart
what is the table structure of vwaccsummary?
2016-04-28 03:27:41
dragonheart
SELECT SUM(credit) - SUM(debig) as balance FROM ....
2016-04-28 03:27:48
dragonheart
woudl eb simplier
2016-04-28 03:28:14
dragonheart
its usually easier to wrap these kinds of queries in application code rather than SQL FUNCTIONs
2016-04-28 03:28:33
jonrwads
See that is where it gets really ugly and I don't think the actual query is the problem as it runs really quickly on it's own and on first run. If I run just the query it return always in under a second. It is only when I call it inside the function
2016-04-28 03:29:05
jonrwads
And only after the first call
2016-04-28 03:34:38
dragonheart
probably after the first call it works out its query the whole table search. SHOW CREATE TABLE vmaccsummary; and EXPLAIN {sqlquery}
2016-04-28 03:46:49
jonrwads
I found this small parameter that solved it stored_program_cache = 0 it turns out that query cache and this cache are separate
2016-04-28 04:03:00
dragonheart
jonrwads: i could say you are going the wrong track, however since you really don't want to listen/communicate to questions asked it would be pointless.
2016-04-28 09:13:28
montywi
serg: I am now ready to push into 10.1; However this needs to be merged quickly into 10.2
2016-04-28 09:13:36
montywi
Do you plan to do a merge today or is it ok if I do it ?
2016-04-28 09:14:08
serg
I don't plan to do anything 10.2 related this week
2016-04-28 09:14:21
montywi
(as some part of the merge is a bit messy because of conflicting code)
2016-04-28 09:14:32
montywi
ok if I do it ?
2016-04-28 09:14:53
serg
sure
2016-04-28 09:14:54
montywi
I assume it's just to do 'git merge 10.1' in a 10.2 branch ?
2016-04-28 09:20:17
serg
yes
2016-04-28 09:38:25
jkavalik
I have a server (MariaDB 5.5.39) where Questions = 41B, Queries = 57B and Handler_commit = 75B - how can there be more (internal) commits than actual statements (including selects)?
2016-04-28 09:38:38
jkavalik
*5.5.33
2016-04-28 09:52:23
montywi
jkavalik: the commit count is per storage engine. Are you using more than one storage engine for some of your queries?
2016-04-28 09:52:42
montywi
and are you checking the global variables?
2016-04-28 09:54:06
jkavalik
montywi, its from "show global status;", every table should be InnoDB - it is one node of 2-node Galera cluster. Might Galera replication count as another commit?
2016-04-28 10:50:45
qwebirc93589
I'm wondering if someone could help me get MariaDB working again on my Debian machine; I removed my binary logs for mariaDB by using 'rm -rf', which caused mariaDB to stop working. I tried to fix it by reinstalling it using apt-get, but I'm unable since the install- and uninstall script checks if there is any mysql-instance running on the computer, which there isn't but it's still failing. There's no mysql.sock in /var/run and th
2016-04-28 10:51:12
qwebirc93589
... All of the dependencies for mariaDB is reinstalled successfully, but mariadb-server-core-10.0 fails, as attached: http://pastebin.com/raw/xuCRxmDa
2016-04-28 10:51:17
jkavalik
qwebirc93589, locate and check the mysqld log
2016-04-28 10:52:00
jkavalik
ah, error during the reinstall..
2016-04-28 10:52:41
qwebirc93589
Hi jkavalik, thanks! Both /var/log/mysql.err and /var/log/mysql.log is empty.
2016-04-28 10:53:42
jkavalik
qwebirc93589, "warning: package is in a very bad inconsistent state; you should reinstall it before attempting a removal" did you try to reinstall it?
2016-04-28 10:55:14
qwebirc93589
Yes, if I try to install it, it fails with the same errors (line 7 - 15)
2016-04-28 10:55:49
jkavalik
qwebirc93589, do you have aptitude installed? I have some experience with it being able to better fix some problems, but no idea if it might help in this case
2016-04-28 10:56:05
jkavalik
qwebirc93589, https://www.maxoberberger.net/blog/2014/07/manually-remove-broken-ubuntu-package.html
2016-04-28 11:02:14
qwebirc93589
jkavalik, Oh, thanks! I actually managed to remove the package with those instructions, but when I try to install it again it fails with this: http://pastebin.com/ZbjG3zXV
2016-04-28 11:03:56
jkavalik
qwebirc93589, "There is a MySQL server running, but we failed in our attempts to stop it." - "ps ax | grep mysqld" to see if it is running, then try to kill it and see if it shows again
2016-04-28 11:11:54
qwebirc93589
This is the only hit i get with ps aux, and I guess its the grep command itself: http://pastebin.com/iku8sJPB
2016-04-28 11:12:37
qwebirc93589
jkavalik, I forgot to mention you.
2016-04-28 11:13:13
jkavalik
qwebirc93589, seems so, check if the socket file exists maybe?
2016-04-28 11:14:23
jkavalik
qwebirc93589, maybe try running "sudo killall mysqld" anyway?
2016-04-28 11:14:54
qwebirc93589
jkavalik, nothing related to mysql/mariaDB in /var/run/. "sudo killall mysqld" returns "mysqld: no process found".
2016-04-28 11:17:04
jkavalik
qwebirc93589, weird.. maybe try to remove/purge mysql-common and mariadb-common? it looks like existence of some init files is enough for the error to show up when there is no real process to stop - https://github.com/MariaDB/server/blob/10.1/debian/mariadb-server-10.1.preinst#L25
2016-04-28 11:17:35
jkavalik
does the file "/etc/init.d/mysql" exist?
2016-04-28 11:18:02
qwebirc93589
jkavalik, yes it does.
2016-04-28 11:19:12
jkavalik
try removing the two *-common packages and if it still exist after that, remove it (move it somewhere else to keep backup)
2016-04-28 11:23:01
SeerKan
Hi guys
2016-04-28 11:23:33
SeerKan
When installing mariadb on ubuntu apt will start the server by default. Any way to make it not start on install ?
2016-04-28 11:25:17
qwebirc93589
jkavalik, so I removed the file (/etc/init.d/mysql) and made sure that every package was removed (everything related to mysql and mariadb), then I tried to install it again and got this: http://pastebin.com/Z3jnKKbi
2016-04-28 11:28:03
jkavalik
qwebirc93589, "128 rollback segment(s) are active." - sounds like some logs (and probably data) in your datadir are still there and not in 100% state, do you need to backup the data or could you purge the datadir and refresh from a backup?
2016-04-28 11:31:06
qwebirc93589
jkavalik, the data is not needed at all. It would make sense that it's not in a consistent state since I removed the binary logs. Do you know which dirs do I need to purge?
2016-04-28 11:31:47
jkavalik
qwebirc93589, the datadir.. probably the same dir where the logs were (there should be the ibdata file and dirs for databases)
2016-04-28 11:41:32
qwebirc93589
jkavalik, I removed /var/lib/mysql/, /etc/init.d/mysql and /var/log/mysql/ and tried to reinstall after I had purged all mysql- and mariadb packages, and got this: http://pastebin.com/hazzWL6u
2016-04-28 11:42:20
qwebirc93589
I also checked for the socket and pid and neither was running
2016-04-28 11:43:16
qwebirc93589
Oh, and I also found and removed /var/lib/mysql.backup
2016-04-28 11:44:38
montywi
elenst|vac: ping
2016-04-28 11:46:00
montywi
serg: how to setup wsrep library for running tests?
2016-04-28 11:46:10
serg
what do you mean?
2016-04-28 11:46:22
montywi
where can I get the library ? It's not part of mariadb source tree
2016-04-28 11:46:35
montywi
I want to just to do: mysql-test-run --suite=galera
2016-04-28 11:46:45
montywi
to verify my changes works with galera
2016-04-28 11:46:59
elenst|vac
montywi: yes?
2016-04-28 11:47:20
elenst|vac
the library -- you can download sources and build it
2016-04-28 11:47:29
montywi
elenst|vac: thanks, but serg is helping me
2016-04-28 11:47:34
elenst|vac
ah, ok
2016-04-28 11:47:36
montywi
I just wonder how to run the galera test suite
2016-04-28 11:47:47
elenst|vac
yes, you need the library
2016-04-28 11:47:48
serg
montywi: you need to have galera installed
2016-04-28 11:47:51
montywi
as we don't seem to have documentation for that
2016-04-28 11:48:03
serg
you can get from your distribution, I hope
2016-04-28 11:48:04
montywi
serg: galera is kind of part of 10.2;
2016-04-28 11:48:04
elenst|vac
you don't really have to have it installed, you can just have it built
2016-04-28 11:48:40
jkavalik
qwebirc93589, "128 rollback segment(s) are active." - still looks like there are some data/logs somewhere.. check "sudo locate ibdata1"
2016-04-28 11:48:43
montywi
will the wsrep library work both for 10.1 and 10.2?
2016-04-28 11:49:16
elenst|vac
i think so
2016-04-28 11:49:40
elenst|vac
we have a galera repo btw: http://yum.mariadb.org/galera/repo/
2016-04-28 11:49:42
montywi
serg: it's not in suse
2016-04-28 11:50:00
elenst|vac
montywi: see above, there are opensuse13 packages
2016-04-28 11:51:06
montywi
elenst|vac: thanks. trying that
2016-04-28 11:52:19
montywi
and of course the reprot doesn't work :(
2016-04-28 11:52:27
montywi
sorry, the rpms:
2016-04-28 11:52:44
montywi
libboost_program_options.so.1.53.0()(64bit) is needed by galera-25.3.15-1.sles13.suse13.1.x86_64
2016-04-28 11:52:55
montywi
and suse is now using libboost 1.54
2016-04-28 11:54:00
elenst|vac
oh sorry about that.. i even filed a bug for it: https://jira.mariadb.org/browse/MDEV-9817
2016-04-28 11:54:09
elenst|vac
completely forgot
2016-04-28 11:54:44
elenst|vac
then I guess building from sources is your best bet
2016-04-28 11:54:47
montywi
do you know how to compile wsrep; Done that once a LONG time ago
2016-04-28 11:54:58
elenst|vac
iirc it's fairly straightforward
2016-04-28 11:55:28
montywi
any instructions in kb?
2016-04-28 11:55:38
qwebirc93589
jkavalik, "sudo locate ibdata1" outputs "/var/lib/mysql.backup/ibdata1" and "/var/lib/mysql.backup/ibdata1" and both "/var/lib/mysql.backup" and "/var/lib/mysql" is removed and does not exist.
2016-04-28 11:56:01
montywi
ok, found how to do it on ubuntu. Will try that
2016-04-28 11:56:27
elenst|vac
ok
2016-04-28 11:58:24
jkavalik
qwebirc93589, yep, the locate db is updated only sometimes so might not picked it is removed.. then I am quite puzzled.. sorry
2016-04-28 11:59:17
montywi
catphish: see above. Memory_used is fixed for next MariaDB 10.1 release
2016-04-28 11:59:26
elenst|vac
montywi: after you've built the library, see mysql-test/suite/wsrep/suite.pm for locations where it's expected. Otherwise it might be confusing
2016-04-28 11:59:38
montywi
elenst|vac: thanks!
2016-04-28 12:00:13
montywi
elenst|vac: I will add a comment about this to https://mariadb.com/kb/en/mariadb/building-the-galera-wsrep-package-on-ubuntu-and-debian/
2016-04-28 12:00:22
catphish
montywi: excellent, thanks, i'm moderately convinced that i have a memory leak since memory usage continues to increase despite workload and clients remaining the same, is there anything further i can do to investigate?
2016-04-28 12:01:02
catphish
i intent to upgrade to the latest 10.1 and restart everything in the next 24 hours or so
2016-04-28 12:01:19
qwebirc93589
jkavalik, bummer! Thank you so very much for the help and effort any way!
2016-04-28 12:02:15
montywi
catphish: wait until next binary release, unless of course you want to take a snapshot
2016-04-28 12:03:11
catphish
i'll wait for the binary release, but i the meantime i have no choice but to restart anyway, i just wondered if there was anything else i could to to help debug, or is looking at total_memory the next step?
2016-04-28 12:04:39
montywi
looking at total_memory is probably the best step
2016-04-28 12:04:44
catphish
ok
2016-04-28 12:10:06
montywi
catphish: trying another malloc library is next step
2016-04-28 12:10:16
catphish
ok
2016-04-28 12:34:53
qwebirc93589
jkavalik, I moved /etc/init.d/mysql to /etc/init.d/bak_mysql, reinstalled and reconfigured mariadb-server-10.0 and switched port to 3307 and now it seems to work. Thank you so very much!
2016-04-28 12:35:17
jkavalik
qwebirc93589, congrats
2016-04-28 13:07:56
svoj
spetrunia: ?
2016-04-28 13:16:58
jkavalik
nice, "select .. from t1 where (a, b, c) in (select x, y, z from t2)" uses derived_merge even in 5.5, comes in handy for limiting mysqldump with --where
2016-04-28 13:37:28
Trepliev
Hi everyone. Just upgraded MaxScale from 1.3.0 to 1.4.2 and maxscaleis now unable to start. With example configuration file it will start ok. Version 1.3.0 was ok with my config. (log: http://pastebin.com/cjj6LzJA) (conf: http://pastebin.com/fxDTp6Sd)
2016-04-28 13:37:58
markusjm
Trepliev: we've just learned of this bug and we're aiming to release a fix ASAP
2016-04-28 13:38:12
markusjm
relates to this: https://jira.mariadb.org/browse/MXS-700
2016-04-28 13:39:01
Trepliev
markusjm: ah, ok, thank you very much, i will wait fox fix. Thank you markusjm
2016-04-28 13:48:11
spetrunia
svoj: ?
2016-04-28 13:50:04
svoj
spetrunia: can we have a call?
2016-04-28 13:50:19
spetrunia
svoj: yes, in 3 min
2016-04-28 13:50:29
svoj
spetrunia: skype me when ready please
2016-04-28 13:51:38
montywi
nirbhay: ping
2016-04-28 13:51:49
nirbhay
montywi, pong
2016-04-28 14:19:49
jkavalik
is it possible that using Galera inflates the Handler_commit counter? I just checked our second Galera cluster (10.0 this time) and on the server used for writes it is again greater than both Queries and Questions (all but one table in the "main" database are InnoDB)
2016-04-28 15:13:44
Nakiri
Hello. Could someone help me with Maxscale?
2016-04-28 15:14:06
markusjm
Nakiri: having problems with it?
2016-04-28 15:14:26
Nakiri
What does "No. of hangup events: 71797" mean in the "list epoll" output?
2016-04-28 15:14:35
Nakiri
Sounds like something I'd really not want to have.
2016-04-28 15:15:10
markusjm
those events happen when the client closes the connection before MaxScale does
2016-04-28 15:15:27
markusjm
I'll have to double check but I think it's somewhat harmless
2016-04-28 15:16:49
Nakiri
Alright, thank you so much!
2016-04-28 15:17:13
markusjm
it counts the number of EPOLL_HUP and EPOLL_RDHUP events
2016-04-28 15:22:15
montywi
svoj: you added 'enable connection log by default'
2016-04-28 15:22:28
montywi
however you forgot to update all galera tests...
2016-04-28 15:27:18
montywi
and apparently also a lot of other tests :(
2016-04-28 15:28:34
svoj
montywi: that could easily happen, I updated only those tests that were covered by buildbot
2016-04-28 15:28:46
montywi
i just did a full test and got warnings
2016-04-28 15:29:02
montywi
maybe these are tests from 10.1 that I have merged in
2016-04-28 15:33:40
nirbhay
svoj, FYI, I will update the galera tests.
2016-04-28 15:34:15
svoj
nirbhay: thanks!
2016-04-28 15:34:29
nirbhay
np
2016-04-28 15:43:02
catphish
i've got a replication slave that seems to be making a huge number of disk writes per second (far more than the master), is this likely to be due to some kind of fragmentation
2016-04-28 15:43:44
catphish
it appears to be constantly in a state of "Unlocking tables"
2016-04-28 15:51:23
svoj
dbart: Hi! Was p8-rhel6 upgraded recently?
2016-04-28 16:01:44
catphish
i wonder if my issue is row based replication causing a lot more load than queries on the master
2016-04-28 16:07:09
dbart
svoj: yes, I had to rebuild it yesterday
2016-04-28 16:09:43
svoj
dbart: you forgot to run ldconfig from ATC :)
2016-04-28 16:10:02
dbart
hmmm... I thought I did, I'll do it again...
2016-04-28 16:10:22
svoj
dbart: ...after installing jemalloc
2016-04-28 16:10:55
dbart
ahhh, that's probably where I went wrong :)
2016-04-28 16:11:43
svoj
:)
2016-04-28 16:11:46
dbart
svoj: ok, done
2016-04-28 16:11:51
svoj
dbart: thanks a ton!
2016-04-28 16:12:00
dbart
you're welcome!
2016-04-28 16:21:04
catphish
so, my master is generating 45 disk writes per second, but my slave, which is doing nothing but row based replicaton from the master is generating 220 writes per second, is there some obvious reason for the discrepency?
2016-04-28 16:23:24
montywi
in theory, you should get same number of writes/second ; Less on the slave is binary logging is not enabled
2016-04-28 16:29:29
catphish
i'm inclined to agree
2016-04-28 16:32:25
catphish
it may be related to my backup software snapshots, i'll test again when i don't have an active snapshot
2016-04-28 18:18:40
Lazylegs
Hello, question about the column store
2016-04-28 18:19:07
Lazylegs
I have not watched the webinar yet. But which version might include it?
2016-04-28 18:20:01
Lazylegs
We start building new application which will have extensive logging. TokuDB could do the job but I guess it would lack the versatility of it
2016-04-28 18:20:30
Lazylegs
If I can get my boss around to get dedi/vps that we would manage ourself
2016-04-28 18:23:15
Lazylegs
Trying to keep us away from forced services that will run mysql 5.5. or 5.6. Just cause of plenty of things, one is the engines that MariaDB has
2016-04-28 19:19:55
montywi
Lazylegs: Columnstore will be based on 10.1
2016-04-28 19:26:10
F2Knight
Q: I just tried to move the matadir but am greeted with access errors, yet all the permissions are the same. (cp -RPuv /var/lib/mysql /newdir/) and I am running ubuntu 14.04... there appears to be no app armor preventing access... baffled.
2016-04-28 19:26:45
F2Knight
http://pastebin.com/RsP9s7fu
2016-04-28 19:30:05
catphish
so, my slave is definitely creating a very different disk IO pattern than my master, they write a comparable amount of bytes, but the slave created 5x as many write operations to achieve it
2016-04-28 19:30:48
F2Knight
http://pastebin.com/mzqkc3sd <- full error mesage
2016-04-28 19:31:43
catphish
http://paste.codebasehq.com/pastes/tkdd8b02wz60xuw943
2016-04-28 19:32:26
grknight
F2Knight: the user mysqld is running has must have full access to /home/mysql, as well as entry access (x) to /home
2016-04-28 19:32:58
F2Knight
I have chown -R mysql:mysql /home/mysql
2016-04-28 19:33:28
grknight
F2Knight: `ls -ld /home /home/mysql`
2016-04-28 19:34:20
F2Knight
root@pbx-recordings:~# ls -ld /home /home/mysql
2016-04-28 19:34:20
F2Knight
drwxr-xr-x 6 root root 4096 Apr 28 10:18 /home
2016-04-28 19:34:22
F2Knight
drwxr-xr-x 4 mysql mysql 4096 Apr 28 11:15 /home/mysql
2016-04-28 19:34:43
F2Knight
sorry i should of pb that
2016-04-28 19:35:52
x86iac
a/ca
2016-04-28 19:36:06
grknight
it should work unless it is not starting as the mysql user.. unless selinux or apparmor or some other "security" is blocking it
2016-04-28 19:37:12
F2Knight
apparomor file is empty (with note from maria saying why) and if I change the datadir back to the original it starts right up... I have been trying all morning.
2016-04-28 19:37:34
F2Knight
there is no other 'security' its a vanilla install of ubuntu.
2016-04-28 19:37:59
F2Knight
it should be noted that MySQL had no problem being moved. but maria does for some reason
2016-04-28 19:40:07
x86iac
didn't know selinux still existed?
2016-04-28 19:41:18
thumbs
x86iac: why wouldn't it exist?
2016-04-28 19:41:27
x86iac
are you using mysqld --verbose --user=mysql
2016-04-28 19:41:29
thumbs
it's widely used in rpm-based distros.
2016-04-28 19:41:55
x86iac
cause the nsa isn't even using the mess snowden adventure proved that
2016-04-28 19:42:09
x86iac
it might be in distros but it shouldn't be
2016-04-28 19:42:18
thumbs
x86iac: don't be silly.
2016-04-28 19:42:44
x86iac
lol you have no clue what you are saying do you? do a simple google
2016-04-28 19:44:22
thumbs
selinux is still a powerful security layer, and has proved to be useful for many production servers.
2016-04-28 19:46:31
x86iac
sad that you believe that but hey - i'm gonna assume you can say with a straight face that windows computers have never had any security issues :)
2016-04-28 19:47:12
thumbs
x86iac: what the heck are you babbling about? I never claimed that selinux was bullet proof! It's just another (useful) security layer.
2016-04-28 19:47:19
F2Knight
WOW Really https://www.reddit.com/r/linuxadmin/comments/2ebhpf/adventures_in_moving_mariadb_data_folder/
2016-04-28 19:48:56
x86iac
thumbs didn't you get kicked out of #httpd for you bs?
2016-04-28 19:49:27
thumbs
x86iac: perhaps it's time for you to stop with the blatant trolling and nonsense.
2016-04-28 19:49:39
x86iac
they deoped you in a flash didn't they
2016-04-28 19:50:01
thumbs
focus on actually helping that user instead.
2016-04-28 19:50:19
x86iac
lol obsfucate and avoid the fact
2016-04-28 19:51:12
F2Knight
boys are we done with the little mine is bigger game?
2016-04-28 19:52:03
thumbs
F2Knight: indeed, I should have ignored him from the get go
2016-04-28 19:52:06
grknight
F2Knight: that's funny. i've never had this problem running multiple instances with different data directories. but i don't use mysqld_safe
2016-04-28 19:52:55
x86iac
thumbs you misinformation got you canned in #httpd you should have learned and done some basic homework instead
2016-04-28 19:53:12
F2Knight
grknight Luckily this is a new install and no data to actually 'move' but I was wanting to use maria for a number of reasons. just I need the data to be on a second drive. If I can not resolve this... then maria will sadly not be an option.
2016-04-28 19:53:24
F2Knight
I did try this with just stock mysql and it worked perfectly first time
2016-04-28 19:53:54
thumbs
F2Knight: namei -mo /home/mysql might be handy
2016-04-28 19:54:48
thumbs
F2Knight: I can't see your full error message, but if the permissions are fine, then a (13) is always caused by another security layyer like selinux/apparmor
2016-04-28 19:54:50
F2Knight
sure same as before
2016-04-28 19:55:03
grknight
F2Knight: btw, what do you have for basedir and/or datadir in the configuration? `mysqld --help --verbose |grep 'datadir\s'`
2016-04-28 19:55:47
F2Knight
grknight let me check, I did not change anything but the datadir. not sure if base is even in the config
2016-04-28 19:56:53
F2Knight
intersting output http://pastebin.com/xpx0Q1Tg
2016-04-28 19:57:22
F2Knight
btw grknight thanks for providing actual commands, its a nice change from some other rooms
2016-04-28 19:57:35
grknight
i use the `` as delimiters to run.. not meant to include them on the command line
2016-04-28 19:58:28
F2Knight
Ohh... um okay ... 160428 11:58:01 [Note] Plugin 'FEEDBACK' is disabled.
2016-04-28 19:58:29
F2Knight
160428 11:58:01 [ERROR] mysqld: Can't find file: './mysql/plugin.frm' (errno: 13)
2016-04-28 19:58:30
F2Knight
160428 11:58:01 [Warning] Could not open mysql.plugin table. Some options may be missing from the help text
2016-04-28 19:58:30
F2Knight
datadir /home/mysql/
2016-04-28 19:58:42
F2Knight
http://pastebin.com/XxTBDx19
2016-04-28 19:59:01
F2Knight
F2Knight need to find a pastebin plugin for textual
2016-04-28 19:59:20
grknight
grknight uses wgetpaste which accepts commands
2016-04-28 20:00:34
thumbs
F2Knight: sprunge is awesome
2016-04-28 20:00:46
thumbs
F2Knight: <command> | curl -F 'sprunge=<-' http://sprunge.us
2016-04-28 20:00:55
F2Knight
your using a terminal to irc grknight ? I am being a little lazy and on a mac :)
2016-04-28 20:01:03
thumbs
use cat and such
2016-04-28 20:01:30
F2Knight
well another day for that for now I would like to get MariaDB working ... / :(
2016-04-28 20:02:10
x86iac
you are using mac as a server ? lol
2016-04-28 20:02:59
F2Knight
x86iac no just as a glorified terminal.
2016-04-28 20:03:07
x86iac
can't stand much more of this ... i guess thumbs can help you with selinux on your mac lol
2016-04-28 20:03:17
F2Knight
grknight my.cnf has basedir=/usr
2016-04-28 20:03:49
F2Knight
no selinux on mac where you got that from i have no idea
2016-04-28 20:04:10
thumbs
F2Knight: I would just ignore that user - they seem to be intent on trolling
2016-04-28 20:04:28
thumbs
anyway, I can't see the full error, but I presume that grknight has
2016-04-28 20:04:47
F2Knight
i pb it
2016-04-28 20:04:58
thumbs
right. pastebin.com times out here.
2016-04-28 20:05:25
F2Knight
oh...
2016-04-28 20:05:49
F2Knight
not sure where else to post it
2016-04-28 20:06:55
thumbs
but grknight already commented on it, so let's wait for him
2016-04-28 20:06:59
grknight
thumbs: i still like running `wgetpaste -c 'mysqld --help --verbose |grep \'datadir\s\''
2016-04-28 20:07:18
thumbs
grknight: a good option too
2016-04-28 20:09:14
grknight
F2Knight: i don't see any issue.. unless mysqld_safe is doing something funky. does `mysqld --user=mysql` start up normally?
2016-04-28 20:09:40
thumbs
just for kicks, does it work if you use another path (not in /home) ?
2016-04-28 20:09:52
F2Knight
160428 12:09:34 [Note] mysqld (mysqld 5.5.47-MariaDB-1ubuntu0.14.04.1) starting as process 4442 ...
2016-04-28 20:09:52
F2Knight
160428 12:09:34 [Warning] Can't create test file /home/mysql/pbx-recordings.lower-test
2016-04-28 20:10:05
F2Knight
thumbs let me try on a different directory not on a mounted drive
2016-04-28 20:10:18
thumbs
F2Knight: yes, your issue is definitely with the mount options.
2016-04-28 20:10:53
F2Knight
thumbs I doubt that as I have other things working here... actually the drive is mounted and servers as a NFS share
2016-04-28 20:11:33
grknight
NFS.. that's an issue
2016-04-28 20:11:41
thumbs
quite.
2016-04-28 20:11:51
thumbs
don't use NFS for the datadir
2016-04-28 20:12:31
F2Knight
not its not..
2016-04-28 20:12:51
thumbs
NFS mounts add tons of complications that make it harder to use a datadir, and not to mention ensure the consistency of your data
2016-04-28 20:13:04
F2Knight
the Disc has a dir that is being used for NFS ... that disc will have a different dir used for mysql not shared
2016-04-28 20:13:24
F2Knight
but the disc is NOT mounted as NFS it is being SHARED over NFS
2016-04-28 20:13:48
F2Knight
but to rule it out i made a new dir /mysql
2016-04-28 20:13:51
F2Knight
and same thing
2016-04-28 20:19:29
thumbs
well, I don't have a mac to test, unfortunately
2016-04-28 20:20:20
F2Knight
UGH.
2016-04-28 20:20:23
F2Knight
no one listens
2016-04-28 20:20:29
F2Knight
this is on an Ubuntu 14.04 system
2016-04-28 20:20:40
thumbs
or, sorry
2016-04-28 20:20:44
thumbs
I don't have ubuntu either.
2016-04-28 20:21:06
thumbs
(I was helping a mac user and crossed some threads)
2016-04-28 20:21:26
thumbs
if you're using ubuntu, check apparmor's logs.
2016-04-28 20:21:53
F2Knight
nothing app armor is not running
2016-04-28 20:21:58
F2Knight
again already covered that too
2016-04-28 20:42:35
F2Knight
GOTIT@! WTF @#$%@$ http://pastebin.com/9g3w8DS2
2016-04-28 20:42:55
thumbs
F2Knight: what was it?
2016-04-28 20:43:12
F2Knight
I want to say nothing ... but ...
2016-04-28 20:43:42
F2Knight
apparently even though app armor was removed... it was still cached ? showed up with the old mysql directories in memory but not in the config.
2016-04-28 20:44:11
thumbs
ah, so apparmor was causing the error.
2016-04-28 20:44:17
F2Knight
sense its a drop in replacement all the files are the same... yes.
2016-04-28 20:45:03
F2Knight
it highlights a bug to me... that the installer doesn't refresh or check the app armor status so having an existing in memory config made it not work.
2016-04-28 20:45:24
F2Knight
I ended up flushing all the apparomore profiles. then moved it and it worked
2016-04-28 21:29:29
catphish
would anybody be able to help explain why my slave is creating a very different disk IO pattern than my master, they write a comparable amount of bytes, but the slave creates 5x as many write operations to achieve it, i suspect it's related to row based replication, but i don't understand it well
2016-04-28 21:35:41
serg
svoj: could you explain why you've suppressed warnings after all?
2016-04-28 21:38:23
serg
ok, email
2016-04-28 22:27:10
SilentNinja1
Hello
2016-04-28 22:28:27
SilentNinja1
I'm new to MariaDB, have some questions on performance.
2016-04-28 22:29:24
SilentNinja1
MariaDB can better handle High concurrency issues, especially heavy updates/delete/inserts on same tables?