CockroachDB beta-20161013

https://jepsen.io/analyses/cockroachdb-beta-20161013

by aphyr • 9 years ago

309 202 9 years ago

BrentOzar2• 9 years ago

"For instance, on a cluster of five m3.large nodes, an even mixture of processes performing single-row inserts and selects over a few hundred rows pushed ~40 inserts and ~20 reads per second (steady-state)."

Wow, that's...not fast.

aphyr8• 9 years ago

I just want to emphasize that:

a.) this is a beta product where the team has been focusing on completeness and correctness before performance

b.) I was testing Cockroach from August through October--Cockroach Labs has put a good deal of work into performance in the last three months, but that work hasn't been merged into mainline yet. I expect things should be a good deal faster by 1.0.

c.) Jepsen does some intentionally pathological things to databases. For instance, we choose extremely small Cockroach shard sizes to force constant rebalancing of the keyspace under inserts. Don't treat this as representative of a production workload.

qaq3• 9 years ago

They need to be at least 1000x to 10,000x faster on the read side though

sametmax9• 9 years ago

Even if they manage to multiply this by 100 on the final release, it's still way weaker than a regular sql db. I hope they have another selling point than performance.

tormeh• 9 years ago

It's named cockroachDB...

AsyncAwait5• 9 years ago

Horizontal scaleability might be one.

flyingramen2• 9 years ago

There is a better response to this up in this thread (https://news.ycombinator.com/item?id=13661735).

The test is testing the worst-case scenario of everything needing locking. CockroachDB uses an optimistic "locking" approach which makes this bad. But if you're use-case is strictly linearizable high-throughput reads good luck finding anything that is not a single-machine database.

bdarnell3• 9 years ago

(Cockroach Labs CTO here)

Jepsen is essentially a worst-case stress test for a consistent database: all the transactions conflict with each other. CockroachDB's optimistic concurrency control performs worse with this kind of workload than the pessimistic (lock-based) concurrency control seen in most non-distributed databases. But this high-contention scenario is not the norm for most databases. Most transactions don't conflict with each other and can proceed without waiting. Without contention, CockroachDB can serve thousands of reads or writes per second per node (and depending on access patterns, can scale linearly with the number of nodes).

And of course, we're continually working on performance (it's our main focus as we work towards 1.0), and things will get better on both the high- and low-contention scenarios. Just today we're about to land a major improvement for many high-contention tests, although it didn't speed up the jepsen tests as much as we had hoped.

bfrog• 9 years ago

Couldn't this be a problem for application authors and a potential source of DoS activities? It seems this should be a big red warning flag somewhere.

andymans• 9 years ago

I can verify.

I've been testing with node.js's pg client (not even using the 'native' option) and on 1 node I get > 1,000 single inserts /sec. I'm seeing this scale linearly too. NB I'm testing on a pretty overloaded macbook with only 8gb RAM. On a decent-sized cluster with production-grade hardware and network, performance is really unlikely to be a bottleneck. Also, it'll significantly improve during this year, one imagines.

tveita• 9 years ago

I wonder how the clock skew bounds affect performance. Would a more accurate clock increase throughput?

aphyr8• 9 years ago

The clock skew limit controls the maximum latency on reads--so yeah, if you're doing reads that contend with writes, improving clock resolution improves latency. Not entiiirely sure about the throughput impact though.

bluejekyll• 9 years ago

Should the clock offset between two nodes grow too large, transactions will no longer be consistent and all bets are off.

A concern I have is that there could be datacorruption in this context, no? It sounds like on a per cluster basis, i.e. Same Raft cluster, there is a strict serializability guarantee. But without linearalizability system wide, you could end up with data corruption on multi-keyspace writes.

mixedCase2• 9 years ago

If I understood correctly, nodes get killed if they drift outside the max time offset tolerance.

aphyr8• 9 years ago

You're correct--large clock anomalies (or VM pauses, message delays, etc) can cause serializability anomalies. I don't exactly know what will happen, because our tests explicitly focused on staying within clock skew bounds.

To speculate: if clocks do go out of bounds, I don't... think... you can violate single-key serializability, because that goes through a single Raft cluster, but you can definitely see multi-key consistency anomalies. I think you could also get stale single-key reads, because the read path IIRC relies on a time-based lease. Might be other interesting issues if Cockroach is shuffling data from one Raft ensemble to another--I think the range info itself might have a time-based lease that could go stale as well. Perhaps a Cockroach Labs engineer could comment here?

Note that the time window for these anomalies is limited to a few seconds-nodes will kill themselves when they detect clock skew.

bdarnell3• 9 years ago

(Cockroach Labs CTO) The major issue with clock offsets is stale reads due to the time-based read lease. Pretty much everything else is independent of the clocks. Stale reads are still enough to get you in trouble (some of the jepsen tests will show serializability violations if you disable the clock-offset suicide and mess with the clocks enough), but it's tricky to actually get bad data written into the database, as opposed to returning a stale-but-consistent view of the data.

exhilaration• 9 years ago

Like Spanner, Cockroach’s correctness depends on the strength of its clocks... Unlike Spanner, CockroachDB users are likely deploying on commodity hardware or the cloud, without GPS and atomic clocks for reference.

I'm curious, do any of the cloud providers offer instances with high-precision clocks? Maybe this is something one of the smaller cloud providers might want to offer to differentiate themselves - instances with GSP clocks.

Already__Taken• 9 years ago

Why can't these datacenters be using a high precision gps source for their NTP source in the building?

tdrd• 9 years ago

[Cockroach Labs employee]

In addition to being extremely precise, TrueTime is explicit about error in its measurements.

In other words, a call to TrueTime returns a lower and upper bound for the current time, which is important for systems like Spanner or CockroachDB when operations must be linearized.

In effect, Spanner sleeps on each operation for a duration of [latest_upper_bound - current_lower_bound] to ensure that operations are strictly ordered.

Your suggestion might improve typical clock synchronization, but still leaves defining the critical offset threshold to the operator, and importantly leaves little choice for software like CockroachDB other than to crash when it detects that the invariant is violated.

shin_lao• 9 years ago

NTP isn't enough, you need PTP.

mikecb• 9 years ago

The spanner announcement mentioned offering direct access to the TrueTime api sometime in the not too distant future.

Edit: Quote from Brewer's post: "Next steps For a significantly deeper dive into the details, see the white paper also released today. It covers Spanner, consistency and availability in depth (including new data). It also looks at the role played by Google’s TrueTime system, which provides a globally synchronized clock. We intend to release TrueTime for direct use by Cloud customers in the future."

aphyr8• 9 years ago

Bingo. I've been hoping Google (or, heck, any colo provider) would offer this for a few years. Semi-synchronous clocks enable significant performance optimizations!

BrentOzar2• 9 years ago

I'm curious, do any of the cloud providers offer instances with high-precision clocks?

Essentially that's what Google's doing by selling Spanner, but you don't get access to the clocks directly. I'd guess that as they get more experience selling Spanner, and they bring the price of the hardware down, they'll rent those directly too.

nindalf• 9 years ago

They're planning to offer direct access to the TrueTime API soon according to the post by Eric Brewer.

tyingq• 9 years ago

I'm curious, do any of the cloud providers offer instances with high-precision clocks?

That's a great question.

Could be a little tricky with AWS for example. It's typical to run a database on a VPC (virtual private cloud), since that layer of the stack doesn't need to be exposed to the Internet. Unfortunately, that means the servers in the VPC can't get to the ⁎.amazon.pool.ntp.org servers.

Skimming around, I don't see any of the cloud providers, VPS providers, or dedicated hosting companies offering anything special, like a local stratum 1 time source for each region/location.

sroussey• 9 years ago

How does this compare to gundb.io?

010a• 9 years ago

Gundb is garbage.

aphyr8• 9 years ago

To be a tad more concrete: GunDB can run in the browser as well as on servers; CockroachDB is a traditional, client/server database. GunDB is aiming for eventual consistency, but when I looked a year ago, it seemed like the algorithm they chose wasn't actually convergent--nodes could diverge because non-commutative updates could be applied in different orders on different nodes. CockroachDB, by contrast, is a serializable, single-key linearizable store. I don't think GunDB offers transparent sharding or SQL, both of which are important aspects of CockroachDB.

Both depend on wall clocks.

vonklaus2• 9 years ago

Wow. Probably fundraising and getting out front of googles new consumer spanner service. I just started experimenting w/ cockroachdb and it sounds/seems great. They def are campaigning hard, landing a Jepsen test and a lot of posts/articles last couple.

I am pulling for them over G on principle. I infer they know people at Stripe where the Rethink team & Aphyr are so hopefully they can learn from them and build an awesome product. G has lost a lot of battles on UX, Docs, attitude and just general product. They have won a lot of battles on sheer size & resources. Be nice to have this stick around.

aphyr8• 9 years ago

Just as a headsup, I left Stripe in late 2015 to run Jepsen as an independent consultancy.

baq• 9 years ago

i wanted to thank you for your work, your blog is singular on the web. chapeau bas.

seahckr• 9 years ago

big fan of your work! Thank you

vonklaus2• 9 years ago

Hey man, didn't realize you left Stripe but did sort of notice Jepsen getting more serious. Jepsen has phenomenal reach:

Interesting and hilarious coverage of technical complexity which assumes just enough knowledge to engage a keen novice while seemingly engaging seasoned experts in the field.

Thanks for that. Your entertainment firm is a bit out if my wheelhouse, but it's such a crowded space and you end up getting really tied down. Maybe get acquihired for an Iditarod team, but I reckon this is more scaleable.

Posts are great. Tganks

will_pseudonym• 9 years ago

CockroachDB might be the worst product name I've come across. I get it; cockroaches are extremely hardy, but they're also much more strongly associated with uncleanliness and disease.

EDIT: Also, cockroaches are BUGS! Bugs cause problems in computer systems.

artursapek2• 9 years ago

I think it's the most memorable database name I've seen. So at least it has that going for it.

atmosx2• 9 years ago

I'm not a native English speaker and have troubles pronouncing "cockroach". In my view the naming is rather unfortunate.

cocktailpeanuts• 9 years ago

People will remember if the product is good so if you're confident in your technology you don't need any gimmicky tactics.

They're just hurting themselves for no good reason other than feeling good about their bad sense of humor most people don't sympathize with.

Imagine any IT department in a large company, when they're deciding which technology to use, this definitely matters. I mean, funny names like "mongodb" is alright, but the first image when people hear about "cockroachdb" is not a good picture. And first impression matters a lot. It's marketing 101.

Just think about real-life scenario of how you would recommend to a co-worker:

"Hey we should definitely use this new database. It's called couchdb!" vs "Hey we should definitely use this new database, it's called cockroachdb!"

Or,

"Hey we should check out this awesome way of representing API. It's called JSON-API!" vs "Hey we should check out this awesome way of representing API. It's called Swagger!"

zzzcpan4• 9 years ago

Agreed, they can change it if they ever get big enough. But before that more memorable brand is going to make it feel more trustworthy than the competition.

radub• 9 years ago

http://morgana249.blogspot.com/2014/10/images-of-most-beauti...

MikeAMartinez• 9 years ago

I always thought MongoDB was a pretty horrendous, semi-racist name and they ended up dominating, for better or worse, the NoSQL market. Of course, I quickly found out that Mongo was meant to be related to term 'Humongous' and not 'Mongoloid'.

niyazpk• 9 years ago

Created an issue here: https://github.com/cockroachdb/cockroach/issues/13634

Scaevolus• 9 years ago

The founders of Cockroach Labs are also the creators of GIMP.

WhitneyLand3• 9 years ago

Could the name actually slow the project's momentum?

I only recently read some cool things about CDB and wondered if I could have been subconsciously skipping articles about it for deep seeded reasons like http://bbc.com/future/story/20140918-the-reality-about-roach...

Of course anything good may eventually rise to success on its merits, but in case I'm not the only one with a subconscious aversion maybe the info below will help put some people one less click away. How about SurvivalDB? :)

CDB is a distributed SQL database built on a transactional and strongly-consistent key-value store. Scales horizontally; survives disk, machine, rack, and datacenter failures with low latency disruption and no intervention; strongly-consistent ACID transactions; SQL API for querying data. Inspired by Google’s Spanner and F1. Completely open source.

https://cockroachlabs.com/docs/frequently-asked-questions.ht...

finder832• 9 years ago

I agree completely. I love reading about databases, but apparently have been skipping reading about CockroachDB for that same reason subconsciously. I took a deep dive the other day and was amazed that I've seen the name dozens of times and have never bothered to ask what was special about it. It really is a fascinating database.

AsyncAwait5• 9 years ago

Well, given the claimed properties of the database, I can see why they named it this way, but it is not the world's most fortunate name, for sure.

zzzcpan4• 9 years ago

No, you have not been skipping it subconsciously or seeing it more often, than others, you just feel like you did. People actually have an availability bias [1] that clouds their judgement about unusual things.

[1] https://en.wikipedia.org/wiki/Availability_heuristic

omginternets• 9 years ago

Same, to the point where I just now realized I know nothing about CockroachDB...

lallysingh2• 9 years ago

I completely wrote it off as a bug database. Only when it came up in conversation did I look it up.

SilasX4• 9 years ago

I had the same experience with Rust, for the same reason.

geofft• 9 years ago

Yeah, I've definitely heard people tell me "It seems cool, but not cool enough for me to get over the subconscious revulsion to the name" when asking about CDB.

I appreciate the name, but I think I'd have an easier time convincing people to use a product of a different name....

digler9992• 9 years ago

to me it's not a deeply-seeded reason, it's just when I hear cutesy-sounding names I just don't take them seriously. They sound like the latest, hipster thing that's going to be in-bloom for a (figurative) week or two until the next cutesy thing steals everyone's attention. Even the language Pony has a mildly off-putting name (to me) despite hearing people rave about it. Couch DB ? What the hell does that mean ?

I admit it's a bias of mine and I could be losing out on learning amazing tools. But even non-profit, standardized languages and tools still have a brand name and image.

WhitneyLand3• 9 years ago

See your point, but maybe there's a middle ground between hipster and bug like?

huehehue2• 9 years ago

Ah yes, let's not forget the solid branding behind some of Adobe's work:

https://pig.apache.org/

https://hive.apache.org/ (Look at the anguish in this creature's face)

http://hadoop.apache.org/

SilasX4• 9 years ago

I know that was rhetorical, but Couch is connotationally neutral AFAIK, and comes from "cluster of unreliable commodity hardware".[1]

At least, it's miles ahead of Cockroach or Rust or even Mongo.

[1] https://en.wikipedia.org/wiki/CouchDB#History

justicezyx5• 9 years ago

If CDB just wants to be a niche, live in a corner in the shadow of Cloud Spanner or other similar big players, the name is not critical.

Otherwise, its a realistic issue has to be addressed.

OpenDrapery• 9 years ago

I'm wondering if all the people who can't get over the name CockroachDB are also people who insist that professionals should dress the part. Can't take someone seriously if they don't have their shirt tucked in?

Style over substance.

WhitneyLand3• 9 years ago

Definitely not. The opposite in fact, I don't even require shoes or shirt if you're speaking rationally.

chipotle_coyote• 9 years ago

I'm pretty sure I'm not alone in having a bias against the name "CockroachDB" and being okay with people who don't have their shirt tucked in. I get the comparison you're making, but "untucked shirt" connotes "casual" to me. Roaches connote "health hazard." I'm okay with eating at casual restaurants, but not casual restaurants with roaches.

If you have to explain "we want you to think about survivability, not health violations," then maybe the problem isn't your untucked shirt--it's the big grease stain covering its front. You can insist that's just part of your style all you want, but it's going to be a hard sell.

wstrange• 9 years ago

Maybe they should have called it TardigradeDB

theptip4• 9 years ago

TardigraDB?

SilasX4• 9 years ago

Agree 100%, ditto for suggested alternate name. I'd been doing the same for Rust. It's not a rational reason, sure, but I have to sell adoption to a broader userbase that cares about more that the technical merits.

I don't want to have to make the case for why someone's system needs rust and cockroaches.

JaggerFoo• 9 years ago

"Rust and Cockroaches" is the title of Kenny Chesney's next Country Western song.

AsyncAwait5• 9 years ago

Rust is nowhere near Cockroaches in terms of what it evokes.

logingone• 9 years ago

I have to say, I think that if you have a problem with "Rust" then you lose credibility in your criticism of "Cockroach".

Contrary to most above, I think "Cockroach" is great. It's suitable. Any suit who has a problem with it should get back on their side of the office, and stop sticking their nose in things they know nothing about, especially when so evidently closed minded. There are too many people in that meeting. Hell, even "Google" is a crap name, sounding like a baby word, and "Mongo" can be outright offensive. I expect anyone who has an issue with the name "Cockroach" is probably not going to be too open to the cutting edge anyway, so wasted concern.

Perignon3• 9 years ago

Agree, the name is pretty off putting. Maybe they were trying to be funny? I don't know, but I would recommend changing it.

andrewchambers3• 9 years ago

the level of bikeshedding on the name is crazy. I think it is a good name btw.

renegadesensei3• 9 years ago

Completely agree. I specialize in distributed databases and I have zero interest in even learning about CockroachDB. Could be the greatest technology ever and I will never know. Is that irrational? Sure. But I'm a busy man with plenty of tech to learn about and limited time. I detest roaches and do not want to think about them at all when I am working. Might as well have named it with an ethnic slur.

andrewchambers3• 9 years ago

You seem to specialize in willful ignorance and bike shedding to me.

jdcarter2• 9 years ago

Absolutely the name is holding them back. In my case it prevents me from championing the product within my company. Going to the executives and saying "we're going to put our most precious data into Cockroach" simply won't fly. Any discussion of features will be for naught; I know the CEO will make a "no" decision as soon as he hears the name.

To any Cockroach Labs folks who might read this: please give me (and all of us) a name that I can evangelize without my audience cringing.

corndoge• 9 years ago

If it's not too much trouble you might consider looking for a new job if management is that obtuse.

_xhok4• 9 years ago

Why at your company are people who would evaluate a database by its name rather than its technical merits the ones who decide what database gets used?

influx2• 9 years ago

It's a terrible name, it also has the property that stupidly configured web filters could filter it due to string matching the first part of the word.

It also makes your employees look stupid when they tell their landlords, bank accounts, dates that they work at Cockroach Labs. Good luck explaining that you aren't an exterminator.

irfansharif2• 9 years ago

what's wrong with being an exterminator?

fnord123• 9 years ago

Do you work on influxdb?

crudbug2• 9 years ago

Yes.. RoachDB sounds much better, if they want to continue with the same line of thought.

throwawasiudy6• 9 years ago

their name is the stupidest startup blunder I've ever seen. Enterprise won't go near it, seriously. I cannot tell a client I'm using a "cockroach" database, no fucking way.

If we could go back and rename Windows Hitler would anyone be using it???

justicezyx5• 9 years ago

Imagining introducing CDB in a room full of CXOs from an enterprise customer...

CXO: What CDB stands for? Presenter: Cockroach Database! ...

simplehuman• 9 years ago

If we could go back and rename Windows Hitler would anyone be using it???

That didn't take long. https://en.wikipedia.org/wiki/Godwin%27s_law

SeeDave• 9 years ago

If we could go back and rename Windows Hitler would anyone be using it???

Intensely hyperbolic but... you're not wrong. "Cockroach" is an awful name for a database and I have a hard time envisioning "seen-it-all" enterprise IT managers taking it seriously.

Fiahil• 9 years ago

As a former SC2 Zerg player, I found the name a bit funny, nothing more.

A really bad name, would be, for example, to try calling a news agency "BuzzFeed News".

Birch-san• 9 years ago

I submit "Bunker" as a more approachable name. Not only does it have the connotation of "survives nuclear bombardment", but also suggests "keep your stuff safe here" and "something you could entrust with your life".

adamtulinius• 9 years ago

The name "Bunker" also suggest that the developers are German nazis, since German nazis built a large amount of bunkers in Europe during WW2.

.. so probably not a good name either, at least not in Europe. ;-)

andybons• 9 years ago

Already taken: https://bunkerdb.com/

_mg• 9 years ago

Well, I know I definitely won't forget the name.

ArlenBales• 9 years ago

It might be intentional as reverse psychology. They don't want employees or users who wouldn't use a project because it's named after an unpopular insect. Those users/employees could be fickle, unreasonable and low quality at other aspects of business/development and be a net loss for the company/project in the long term.

manigandham4• 9 years ago

That is a terrible way to run a business.

AsyncAwait5• 9 years ago

I'd be interested in any comparisons of CockroachDB vs TIDB[1], especially when it comes to speed.

1 - https://github.com/pingcap/tidb

placeybordeaux• 9 years ago

Yeah I find it odd that most discussions of CDB don't reference TiDB, but in this case speed is not really the focal point.

ohstopitu• 9 years ago

CDB is a really cool db for the time I've been playing with it (just a few months I think).

That said, I'd really appreciate if they'd launch a hosted DBaaS. By that I mean, more in line with dynamodb or documentdb instead of say MySQL on AWS for example.

I know it's on their timeline, but dbs generally get selected during the start of the projects and unless absolutely necessary, companies don't want to change dbs.

dianasaur323• 9 years ago

CockroachDB employee here. DBaaS is definitely on our roadmap. There are a lot of moving pieces to building out DBaaS, so our focus first is on honing our core product. In the meantime, we are working on making it as easy as possible to deploy CockroachDB across different cloud vendors and with orchestration tools like Kubernetes.

irfansharif2• 9 years ago

I'm disappointed with the commentary here, I for one would think the discussions around building a distributed system with the consistency guarantees claimed here would be far more interesting than the name of said system.

tshannon3• 9 years ago

The name may just be low hanging fruit to comment on if one has nothing else to add to the conversation.

Gigablah• 9 years ago

I think it's hilarious that people are up in arms about the name but apparently something like MongoDB is fine.

justicezyx5• 9 years ago

People care about CDB raises one of the most prominent and easiest fixable problem. I see 0 problem of this.

p4lindromica2• 9 years ago

Just like you think you can fix my shed by painting it yellow, right?

throwawasiudy6• 9 years ago

May I suggest renaming the thing already? 90% of the business world can't touch it with a name like that.

How about Duradero ... Spanish for durable. The name doesn't seem to be taken by any other database type software

dijit9• 9 years ago

I'm going to be the dissenting voice here and say: Use whatever name you like as long as it isn't something like cuntbagDB.

CockroachDB accurately and immediately conveys meaning. It'll survive when nuclear war happens just like a cockroach. And I might work in the games industry but I would never have a hard time selling this. I would basically force them to use it and explain the benefits.

If your company wont adopt technology because it uses a bug as it's name then your company has much greater problems than technology adoption.

throwawasiudy6• 9 years ago

It's not my company or me, it's my clients. Here's a sort of long winded story explaining what I mean that the corporate world can't use anything with the name "cockroach". If you disagree after reading, I'm open for a friendly debate.

Average project, some 60 year old VP signing off on 100k of custom software with 8 pages of lawyer speak contracts is NOT going to allow "his" project to run off of "cockroach database". VP's/CEO's like to take personal ownership of projects they feel will go well, and ones done with us usually do. None of these people want ANY chance of their project getting criticized internally, so nobody will use CockroachDb. It's just not going to happen. Having successful projects is an easy way for these people to gather promotions and respect. Having any part of it do with "cockroach" could result in the whole project being an embarrassment.

Imagine the database goes down one day. One of their low level staff calls us(we have ongoing support contract). Problem gets rapidly escalated to development. One of our lead developers replies to the email thread, that by now has a third of the clients' company on it. "Yeah see the problem was the cockroach database that wasn't supposed to go down needed to be restarted" . Most of the staff think we're joking. Guarantee this becomes a big pun around their office, every time anything breaks in the app they say "maybe the cockroaches on the back end died again".

It's made worse by the fact that issues in software systems are commonly called "bugs". Well, our system doesn't have general bugs anymore, it has "cockroaches". Do you think this is an unlikely scenario? Because it's not. Besides the word "cockroach" this is exactly how urgent issues tend to be reported and fixed. Replace "cockroach" with "Microsoft" and I've probably violated my NDA. After that incident, whoever championed the project, probably our strongest ally in the company, resents us for making such a stupid mistake that resulted in his project becoming a joke. Maybe we aren't as great of a dev shop as he once thought? Maybe it's better to do open bidding next time so whoever wins he can't be blamed directly for what happens.

You could say their company culture sucks, maybe it does? But that doesn't change that I can't use it on any client projects because of the chance we'll lose hundreds of thousands over a stupid name.

There's plenty of other names that convey meaning but don't make lawyers nervous. Your attitude of "oh, well if you can't use it your company culture sucks" is EXACTLY why nobody will rename it and why corporate american won't touch it. Nobody cares about how cool it is to be edgy when millions of dollars are on the line.

If you plan on forcing clients to use it you might as well hold a going away party for any of your clients with more than 5M revenue.

rystsov• 9 years ago

Dura means fool (feminine gender) in Russian :)

crudbug2• 9 years ago

RoachDB !

dijit9• 9 years ago

I wish everyone would stop bikeshedding the damn name.

From my perspective I'm glad something that's trying to do this is coming into existence. a couple years ago I was given the arduous task of ensuring that we never lost data. One of the requirements was that: At any time, any server can (and will) fail.. if you acknowledge that you have data then you MUST never lose it.

You cannot imagine how many database solutions that ruled out. Including basically _all_ noSQL data stores.

We settled on postgresql, because, despite going into toast tables and having to implement sharding on top- it was not only the most sane it could be wrangled into basically doing the right thing in regards to fsyncing and ensuring data wasn't in VFS.

So, Kudos to these guys, who, in a world of people who don't give a shit about consistency (except eventual consistency) are actually trying to further it.

Perignon3• 9 years ago

It's not bikeshedding. Who the fuck wants to work for a company called Cockroach Labs? Seriously, sure your tech friends and recruiters might appreciate it on your resume but anyone outside of tech? Applying for an apartment, mortgage, car, lol good luck. And also not too sure that anyone outside the immediate tech/startup nerd scene would appreciate it other.

That aside why the hell do I want to thing of cockroaches day in and day out when I am using their product? Seriously, is this someone's idea of a bad joke?

The name has got to go.

dijit9• 9 years ago

Nobody is going to give a flying fuck what your company is named on a mortgage application.

Honestly there are companies with _much_ worse names in active use like "Dick & Daughters"[0] or "Ismail ANUS"[1] who even got an appointment to the queen.

Seriously, it's bikeshedding, and the point of the DB is that you don't think about it all day.

I also don't think of women called Cassandra when I use the database of the same name. Bikeshedding. Seriously.

[0] https://beta.companieshouse.gov.uk/company/08610385

[1] https://beta.companieshouse.gov.uk/officers/5-sFnEsibrDnUR2-...

p4lindromica2• 9 years ago

"Who the fuck wants to put their bikes in a magenta shed?"

9935c101ab17a66• 9 years ago

It's not bikeshedding at all. It looks like a good solution, but many people have concerns about an inability to use it because of management/clients and the connotation of the name. That's a legitimate concern. Don't just hand wave it away.

justicezyx5• 9 years ago

Utility wise, naming feedback is much useful.

For any database product, there are plenty of mature and robust technical tools for improving and evaluation. I do not believe the company is less aware of any HN commenter. Therefore any technical feedback only going to have limited impact.

But the naming feedback seriously helps them get an idea how their naming hurts their brand-acceptance.

By all means, that's a valuable feedback.

zzzcpan4• 9 years ago

"We settled on postgresql, because, despite going into toast tables and having to implement sharding on top"

This is a common mistake. The choice doesn't actually guarantee consistency at all. PostgreSQL only guarantees consistency as long as you don't try to communicate with it over a network, which you obviously do if you have shards. To address this problem we have things like 2pc, but they aren't very useful for an RDBMS on their own without the whole infrastructure of a distributed database.

dijit9• 9 years ago

The networking isn't a problem if you only acknowledge a write when the underlying database says so. You could have a write succeed before you crash. But you won't lose an acknowledged write.

throwaway934261• 9 years ago

It would be bikeshedding if we were all, for no good reason and without prompting, suggesting names that we preferred, with no consensus emerging.

Instead what I see in this thread is a clear consensus that the current name is atrocious, even to the point of being detrimental to the long-term success of the project, and should therefore be changed.

Personally, I refuse to use or recommend this product as long as its current name remains. And to those who think I'm being petty, 1) I can afford to be petty given how many database options (NoSQL or otherwise) are out there, and 2) market forces should punish products with names like this, or else we'll eventually find ourselves debating the merits of HerpesDB vs. ManureDB.

ysleepy• 9 years ago

So I assume this means cockroachDB currently (probably) meets its promised consistency levels. How does it compare to the usual default settings of PostgreSQL for example? - I think you get SERIALIZABLE, so the behaviour should be very reasonable.

I understand the txn/s numbers are in a pathological scenario, but will these transactions block the whole db from moving faster or have unrelated transactions still normal throughput?

I'm just glad we get linearizable single keys, I had to switch to HBase because it was the only choice at the time. Cassandra did not even give me read-your-own-writes. HBase also had issues while testing on the same host, delete+write lead to deletes being reordered after the writes because they had the same timestamp.

I'm a bit disappointed from HN in this post, everyone only talks about the name. But 90% probably did not even read the article. Aaanyway, heres my 2cents: https://cr.yp.to/cdb.html is CDB for me :P - so better choose another acronym!

jpgvm2• 9 years ago

I don't believe CockroachDB provides SERIALIZABLE consistency for the entire keyspace, rather for subsets of the keyspace that reside in the same Raft cluster.

In practice this means no cross key serializability without including a serialization token in transactions.

I may be mistaken however but that is my understanding from reading CockroachDB docs and Aphyrs post.

tschottdorf2• 9 years ago

Cockroach Labs employee here. We provide serializable consistency for the whole keyspace. It's just that it's easier when you stay in one Raft ensemble as less can go wrong.

ak4g• 9 years ago

How does it compare to the usual default settings of PostgreSQL for example? - I think you get SERIALIZABLE, so the behaviour should be very reasonable.

Not clear if you're referring to PostgreSQL or CockroachDB, but for the record, the PostgreSQL default transaction isolation level is READ COMMITTED, not SERIALIZABLE.

https://www.postgresql.org/docs/current/static/sql-set-trans...

bdarnell3• 9 years ago

(Cockroach Labs CTO) CockroachDB provides two transaction isolation levels: SERIALIZABLE, which is our default and the highest of the four standard SQL isolation levels, and SNAPSHOT, which is a slightly weaker mode similar to (but not quite the same as) REPEATABLE READ. Unlike most databases which default to lower isolation modes like REPEATABLE READ or READ COMMITTED, we default to our highest setting because we don't think you should have to think about which anomalies are allowed by different modes, and we don't want to trade consistency for performance unless the application opts in.

All transaction interactions are localized to particular keys, so other transactions can proceed normally while there is contention in an unrelated part of the database.

(as for the acronym, we prefer CRDB instead of CDB)

CockroachDB beta-20161013

Related Stories