Kistenmacher: Decoux: oh I don’t have to enable binlogs and then redump the data to fill the binlog?
Conkey: Decoux: also, I may not setup the slave until a few more months
Swader: Is it easy to get the position again later?
Decoux: Scott0_: if you write to the new master, you’ll have to clone the new master to start the slave you can’t possibly store Months of binlogs on the new master
Fluegge: Or just use whatever the current position is when I decide to setup the slave
Anzideo: Binlogs are kind of like round robin storage?
Decoux: Scott0_: the new slave will be so *far* behind that you’ll have to take a new snapshot
Decoux: Scott0_: do you have unlimited storage?
Bendolph: ChrisWJ: try reset slave all
Cova: Decoux: that’s fine, ill snapshot it whenever I need the slave setup. no we do not have unlimited storage
Sieber: Currently its using about 80GB
Decoux: Scott0_: how long can you keep binlogs for?
Goe: Hi. I’m doing some tests with the old mysql-server package in centos6 inside of docker. not sure that matters here but thought I’d throw that out there. For some reason, ‘datadir’ is being reported as: .
Lann: Decoux: until we run out of additional storage I guess, but if the slave is operational, shouldn’t we only need to keep them for enough time for the slave to pick up on the new data?
Sia: I’ve never seen this before. Honestly I’ve never really looked.
Decoux: Scott0_: rotate your binlogs.
Deason: ChrisWJ: so this is some galera node, which will have a master which is giving this error?
Calk: Im guessing the binlogs are configurable as far as the amount of time or amount of space?
Dechamplain: Decoux: is rotation automated by mysql
Gilbertson: Dtrainor: That’s probably the relative path based on the current working dir of the process that started the server.
Decoux: Scott0_: it can be. it’s in the docs.
Franchini: Deason: Correct, this galera node with be a slave to a standalone mysql master.
Podaras: Deason: I’ve set them up before and haven’t had an issue and looking at configs of the other cluster I have, it appears the same. Not sure what I am overlooking.
Sallis: I could see that happening, sure. Though I’m not starting it, just running with –help and –verbose to have mysqld tell me what its perceived configuration is
Deason: ChrisWJ: do you define anything for relay in my.cnf? eg grep -i relay /etc/mysql/my.cnf
Stinser: Dtrainor: It’s more common to use an absolute path.
Decoux: Scott0_: in fact, it *should* *always* be.
Kawczynski: What’s the rule of thumb Decoux? 😛 for buffer pool size and log file size?
Huyck: Dtrainor: Right. That’s probably due to the person that configured the server and startup.
Redfox: Here’s some of the magic that the Dockerfile expects, if it helps give some context: https://github.com/docker-library/mysql/blob/5836bc9af9deb67b68c32bebad09a0f7513da36e/5.5/docker-entrypoint.sh#L11
Grollman: I have increased the memory for the server so I need to update them
Obeso: Sure. datadir is specified in /etc/my.cnf – does mysqld know to first look there by default?
Deason: Scott0_: 70% of total ram or maybe 75%-80% for buffer pool, innodb_log_file_size between 100M and 1024M anything in this range is fine probably
Rodrigus: I can be in any directory and run –help –verbose and mysqld will still list ‘datadir’ as .
Liebowitz: Deason: I tried with defining it and also with not defining it, same issue.
Bernhart: Dtrainor: it’s usually one of the places to look. Yes.
Skarupa: Thanks for confirming
Decoux: Scott0_: it depends. Buffer pool between 50% and 75% of the available RAM
Deason: ChrisWJ: pastebin: show variables like ‘%relay%’;
Rutherford: Currently my log file size was 150MB, how can you tell whether or not this needs increasing?