|
cinci zoo sniper posted:i do recommend options other than postgre only to people who don’t need rdbms recommendations can you rephrase cuz i'm tired and can't understand if you're saying postgres is the default or only for experts or what
|
#
?
Oct 30, 2018 18:39
|
|
|
|
| # ? Apr 24, 2024 17:48 |
|
|
ratbert90 posted:cat bullshit_db.sql |wc -l terrible programmer disclosure - idk what happens after bullshit_db.sql
|
|
#
?
Oct 30, 2018 18:51
|
|
|
NihilCredo posted:can you rephrase cuz i'm tired and can't understand if you're saying postgres is the default or only for experts or what i'm saying that if you have to ask then postgres is the default
|
|
#
?
Oct 30, 2018 18:59
|
|
|
cinci zoo sniper posted:terrible programmer disclosure - idk what happens after bullshit_db.sql The file contains 7,179 lines.
|
|
#
?
Oct 30, 2018 19:01
|
|
|
postgres is so not for experts than even gigantic clown me can manage janitorial on a corporate data warehouse as a "once per two months something happens" thing, where it's usually not something breaking but us just rolling a new postgres version because me or architect found something cool in patchnotes and want to try it btw postgres 11 has cool partitioning, optimizer, and paralellization improvements, as well as addition of a selective jit compiler. a lot optimizer, parallelization, and jit affected stuff is what we all use intesenly on daily basis so the plan right now is to try to roll with the upgrade as soon as next thursday, need to just figure out how i want to go about it since i don't have a second vm to do a cold-start upgrade there
|
|
#
?
Oct 30, 2018 19:05
|
|
|
ratbert90 posted:The file contains 7,179 lines. delet this
|
|
#
?
Oct 30, 2018 19:05
|
|
|
CRIP EATIN BREAD posted:instead of playing catch-up with an inferior database i'd recommend just using postgres because i can't imagine a single use-case where mysql > postgres. maybe it's easier to find mysql dbas than postgres? now mind you i'm not sure i would want a discount mysql dba to be in charge of my db, but i can imagine a manager noticing that one expects less money than the other, and, well...
|
|
#
?
Oct 30, 2018 19:46
|
|
|
Kevin Mitnick P.E. posted:i know how to do a mysql and would need upwards of two hours to transfer those skills to postgres so i could definitely see speccing mysql when idgaf and am not accountable for the success or failure of the product this, but the developer edition of mssql that isn't licensed to host production dbs
|
|
#
?
Oct 30, 2018 19:54
|
|
|
why does postgres use neither mmap nor O_DIRECT in addition to having its own page cache
|
|
#
?
Oct 30, 2018 20:06
|
|
|
a pull request of mine got denied because it wasnt necessary. it was to change all misspelled bandwith methods to bandwidth literally search and replace of new code that will be around for a very long time gently caress me
|
|
#
?
Oct 30, 2018 20:14
|
|
|
karms posted:a pull request of mine got denied because it wasnt necessary. it was to change all misspelled bandwith methods to bandwidth tech debt by design
|
|
#
?
Oct 30, 2018 20:15
|
|
|
karms posted:a pull request of mine got denied because it wasnt necessary. it was to change all misspelled bandwith methods to bandwidth Wait until you start finding variables named bandwidth in the same scope, doing subtly different things.
|
|
#
?
Oct 30, 2018 20:35
|
|
|
cjs: these loving russians. Every single time I pull down a recently-merged stable branch, they introduce 2 regressions for every one fix or feature. It's getting old as gently caress. Next task: convince ceo to drop the russians and hire a couple in-house dudes. I think we're paying at least 2-3 FTE for the russian work, and it's creating more work for us then eliminating at this point.
|
|
#
?
Oct 30, 2018 20:47
|
|
|
Doom Mathematic posted:Wait until you start finding variables named bandwidth in the same scope, doing subtly different things. ah I see we've worked at the same place
|
|
#
?
Oct 30, 2018 20:48
|
|
|
gonadic io posted:tech debt by design we might be the best at this, no lie new non-technical people that worked at our competitors are remarking that we have a sweet setup that's such a step above the competition as is tradition this generates big eye rolls and knowing glances from the technical people, including those who have been here from the beginning and are responsible for over three thirds of this bullshit btw we're hiring!!!
|
|
#
?
Oct 30, 2018 20:53
|
|
|
karms posted:we might be the best at this, no lie AggressivelyStupid posted:ah I see we've worked at the same place e: fortunately new place is fuckin incredible, good shops do exist
|
|
#
?
Oct 30, 2018 20:57
|
|
|
Finster Dexter posted:cjs: these loving russians. Every single time I pull down a recently-merged stable branch, they introduce 2 regressions for every one fix or feature. It's getting old as gently caress. Next task: convince ceo to drop the russians and hire a couple in-house dudes. I think we're paying at least 2-3 FTE for the russian work, and it's creating more work for us then eliminating at this point. hahahaha look forward to your "but they're external so they're not headcount, if they were in house they'd cost 10x as much according to the accountants!" conversation!
|
|
#
?
Oct 30, 2018 21:07
|
|
|
Deep Dish Fuckfest posted:maybe it's easier to find mysql dbas than postgres? now mind you i'm not sure i would want a discount mysql dba to be in charge of my db, but i can imagine a manager noticing that one expects less money than the other, and, well... i highly doubt that anyone who is a "dba" but only knows mysql is going to be useful for anything related to the task
|
|
#
?
Oct 30, 2018 21:12
|
|
|
CRIP EATIN BREAD posted:i highly doubt that anyone who is a "dba" but only knows mysql is going to be useful for anything related to the task we have one of those and 6 months into his employment replication is still not a stable or consistent process even for smallest databases, database access issuance is manual process featuring whitelisting ips of individual employer computers, and host of garbage like 10^9 row unpartitioned tables not being done anything about by him, or anyone else until i finally get through all circles of hell and have a direct conversation with our ceo that his request is impossible because this thing is being ignored 8 months by the gaggle of cretins that are the designers and the maintainers of our production databases. 7 years dba experience. but then again we have 10 year experience database architect who is defeated by concept that gui is optional for doing work so maybe i'm just in the land of cursed database people
|
|
#
?
Oct 30, 2018 21:32
|
|
|
"i have no time for this routine task, i have emergencies unhandled for months" i shriek as i take 24 hours to grant 1 person access to a dozen databases because i manually ssh into each database box to janitor their permissions there and i cannot do that because in half a year i have not configured a database monitoring system more reliable than "we stopped looking at it because it had some weird error messages" so every replication failure or full desynchronisation hits me like american jet bombers iraq
|
|
#
?
Oct 30, 2018 21:34
|
|
|
this guy also has discussed with me rearchitecting our bi (that i sysadmin because i had the idiocy to tell office im not scared from computers) into a distributed system, which is perfectly doable and so i gave him a detailed technical breakdown of the everything, and then yesterday he was actually shocked that i am able to upgrade postgres servers between minor releases????
|
|
#
?
Oct 30, 2018 21:36
|
|
|
"but if i increment the major version every time there's a breaking compatibility change, i'll be at version 100 by the end of the week!"
|
|
#
?
Oct 30, 2018 21:39
|
|
|
CRIP EATIN BREAD posted:"but if i increment the major version every time there's a breaking compatibility change, i'll be at version 100 by the end of the week!" that was in the angle of my technical skills being sufficient to execute such task
|
|
#
?
Oct 30, 2018 21:43
|
|
|
oh jesus
|
|
#
?
Oct 30, 2018 21:44
|
|
|
lol
|
|
#
?
Oct 30, 2018 21:50
|
|
|
CRIP EATIN BREAD posted:oh jesus ikr, im the only system administrator (anything excl. networking) for 3 production systems ran by our team so you'd think, just spitballing here, its safe to assume i can manage minors for modern postgres on modern centos
|
|
#
?
Oct 30, 2018 21:50
|
|
|
cinci zoo sniper posted:"i have no time for this routine task, i have emergencies unhandled for months" i shriek as i take 24 hours to grant 1 person access to a dozen databases because i manually ssh into each database box to janitor their permissions there and i cannot do that because in half a year i have not configured a database monitoring system more reliable than "we stopped looking at it because it had some weird error messages" so every replication failure or full desynchronisation hits me like american jet bombers iraq talk about job security!!!!
|
|
#
?
Oct 30, 2018 21:51
|
|
|
CRIP EATIN BREAD posted:i highly doubt that anyone who is a "dba" but only knows mysql is going to be useful for anything related to the task both are essentially a miserable pile of btrees with SQL on top (ok postgres is a little fancier but still). yeah i think if someone claims to "get" mysql but is helpless with postgres they probably don't have a good grasp of wtf is going on or what they're doing. the fundamentals are the same. different config vars or whatnot should not be a big deal
|
|
#
?
Oct 30, 2018 21:52
|
|
|
sometimes we have little fires around the office where something is stressful or goes wrong and it drives me nuts and then i read poo poo like this and realize i have it way easier than most people
|
|
#
?
Oct 30, 2018 21:53
|
|
|
our current db access granting scheme, as it were: 1) ssh into machine, psql user 2) send to person in skype or email their plain text login/pw then first complaint comes in that user cant connect to the db server host 3) hassle networking people to whitelist specific ip as allowed to connect said db server host (this is done by manually adding a whitelisted client in the router configuration in the office where that person is) then second complaincomes that db server is rejecting connections from that users ip because its not whitelisted 4) ssh into machine 2: vim pg_hba.conf boogaloo
|
|
#
?
Oct 30, 2018 21:55
|
|
|
Your experience has mirrored my own at any company that hires dedicated dbas to do anything.
|
|
#
?
Oct 30, 2018 22:00
|
|
|
thankfully we finally have a cto and i won a fight wherein i no longer had to interact with our foreign database colleagues so my life is simpler these days, i just ping problems to the new guy who was really smug how he knows how to talk to them and everything is smooth guess who’s been having petty slap fights in devops channel every other week for last two months with issues taking days more to be resolved
|
|
#
?
Oct 30, 2018 22:01
|
|
|
Finster Dexter posted:Your experience has mirrored my own at any company that hires dedicated dbas to do anything. our company had developers do everything until i came along and pushed a dba hire through 5 or so months of debate, that included chatting with each development team leader individually to explain to them how dba fits into our organisation, which is textbook definition of data-driven business, and how one can help them and not diminish their authority in some respects same developers also implicitly chose in quorum to have no qa people because the two they hired half a decade ago allegedly were a bad culture fit, so obviously it doesn’t take einstein to conclude that qa is an evil that only serves to erode noble agile traditions
|
|
#
?
Oct 30, 2018 22:04
|
|
|
also sorry for the rant everyone but i literally don’t have a single technically apt colleague i could vent to about this without causing a rumour forest fire
|
|
#
?
Oct 30, 2018 22:05
|
|
|
sometimes I wish I was a total idiot because then I'd be happy just cranking the handle doing stupid bullshit every day like line by line cross referencing data from two systems manually to resolve an exception but instead I flip my poo poo and send emails implying that other teams are morons and asking why their exception process couldn't just, you know, output the data that it failed to match in the exception so nobody had to bother anyway what was I going to say? oh yeah our dbas suck too
|
|
#
?
Oct 30, 2018 22:07
|
|
|
Finster Dexter posted:Your experience has mirrored my own at any company that hires dedicated dbas to do anything. also, last job dbas were cool. i mean they would come with furly brows every time i was convinced i wrote an innovative query to tell me that im an idiot child and suggest a better way, but that's their job imo *excluding that one dba who left the company with inheritance of ~100kloc of undocumented sqls cinci zoo sniper fucked around with this message at 22:10 on Oct 30, 2018 |
|
#
?
Oct 30, 2018 22:07
|
|
|
it was my first job so i was so scared of loving something up that i was running postgres queries without auto-commits and managed to lock up half of important tables on my first week there 
|
|
#
?
Oct 30, 2018 22:08
|
|
|
our quest to become a golang shop is continuing with the assertion that anything on the jvm is just too bloaty for use in our smallish services on k8s that are either crud apps or async don't-give-a-poo poo-about-data stream processors like drat gotta save those jvm startup times when we release, or that docker storage space p.s. speaking of which, this looks pretty good: https://twitter.com/luiyo/status/959808881139601411
|
|
#
?
Oct 30, 2018 22:21
|
|
|
another one we do online payments, we are a php shop, we also have pin terminals and we have only guy who knows c++ and is solely responsible for the code that handles terminal communication. This was fine until it wasn't, so we decided to make this thing less of a headache. The solution we went with? write the thing in php + swoole of course! .... this php abomination is easier to maintain, handles more requests and is way more stable than the c++ code it will replace
|
|
#
?
Oct 30, 2018 22:46
|
|
|
|
| # ? Apr 24, 2024 17:48 |
|
|
karms posted:we are a php shop spookiest thing i've heard all month
|
|
#
?
Oct 30, 2018 22:51
|
|