> "Our cloud offering is “Tiger Cloud.” Our logo stays the same: the tiger, looking forward, focused and fast. Some things do not change. Our open source time-series PostgreSQL extension remains TimescaleDB. Our vector extension is still pgvectorscale."
Small corrections: RocksDB's logo is a cheetah. MariaDB's is a sea lion, which is similar to a seal, but is delightfully relevant to this thread due to sounding more cat-like.
1.x and 2.x are, which is why 3.x reinvents the product around standard tech (true SQL, Apache Arrow). It's hard to ask customers to bet on a database when, to name one reason, its query language has already changed twice.
Every major release of InfluxDB have been a rewrite.
While 3. looks impressive, it seems like most of the interesting features are closed source, so not a 1:1 replacement for version 1.
InfluxDB Edge is open-source, but you need to depend on InfluxDB Community which is free, but closed source, to get things like include functionality like a compactor, which will add capabilities for deletes and re-organizing files to optimize for queries on longer time ranges.
They also need to resurrect all their old 1.* Client libraries for 3.*.
I love InfluxDB, but I’m not hopeful for its future.
InfluxDB Founder & CTO here. We worked hard to support InfluxQL in 3.x and it supports the v1 write API. Admittedly, it will be a migration to move and we haven't yet built the tooling, but we felt it was important to get the 3.0 release out even though we don't have the migration tooling built yet. Our plan is to have that available later this year.
The 2.x to 3.x move is, admittedly, much harder. This is because of the language Flux. We haven't been able to bring that over to 3.x in a way that makes it useful. We actually built a bridge for it in our cloud offering, but our experience is that the performance isn't good enough to be acceptable for customers wanting to upgrade. If they want to make the move, adopting SQL or InfluxQL is likely the only path.
We'll continue to develop 3.x and we'll build more migration tooling over time. I think we can build specialized tooling to help Flux users migrate over to 3.x with query translation tools, but there are more features we need to land in 3.x to enable that first.
We're committed to the technology stack (Apache Arrow & DataFusion) and the 3.x line. We have no plans for another major release. I'll be happy if we end up releasing 3.56.2 8 years from now.
This makes a certain amount of sense because it seems like the actual timescale DB extension/support/etc. they offer is becoming exponentially less important to their company as a result of their pgvectorscale offering. (I'm sure the post says as much.)
I did some work using pgvectorscale and their hosted offering a few months back and the product and the team were a delight to work with. I wish TigerData well.
At first I thought they were sold and the new owner didn't like the original name, but it doesn't seem to be the case. I don't really understand, why would somebody change a recognizable brand.
TimescaleDB will continue to be used to refer to the timeseries postgresql extension. One offering from what they consider to be a larger set of offerings.
Because they don’t want to be pigeonholed as “just time series things”. They continue selling a product called timescale, so I don’t think it’s a loss of brand in much measure.
I talked to the timescale CTO at pg conf a few years ago and asked him what timescale does differently than a standard columnar database that makes it better suited for time oriented data. He said a bunch of things and I said “but columnar databases do those things.” Then he got mad at me.
I guess it’s just another columnar dbms after all?
I'd argue we do okay, but of course it's Clickhouses own benchmark it's hard to outperform them there.
It's also not apples to apples. Clickhouse has much less transactional guarantees and isn't postgres SQL compatible. The great thing about Timescale is that you only need one DB for all your analytics and transactional needs. Combined with pgvector postgres also handles search quite well.
In a way Timescale is just postgres on steroids. Sure if you really know your use-case well, are fine with giving up some postgres nicenes, are willing to learn a new query language and are fine with using and syncing multiple data stores you'll outperform timescale. But I think it is still really cool to see how close you can get with essentially just a better postgres.
"ClickBench evaluates databases using a single table of clickstream data, representative of workloads like web analytics, BI, and log aggregation. It also favors full-table large scans and large-scale aggregations on denormalized data.
Real-time analytics inside applications is different and needs a new benchmark." [0]
This is why we published RTABench. [1]
We believe that it is more representative of real-time analytical workloads.
Do you think all time series databases (like InfluxDB for example) are useless compared to "columnar databases" that "do those things" or just Timescale?
> The majority of workloads on our Cloud product aren’t time-series. Companies are running entire applications on us...
So we are now “TigerData.” We offer the fastest PostgreSQL. ...
Our cloud offering is “Tiger Cloud.” Our logo stays the same: the tiger, looking forward, focused and fast... Our open source time-series PostgreSQL extension remains TimescaleDB. Our vector extension is still pgvectorscale.
Why “Tiger”? The tiger has been our mascot since 2017, symbolizing the speed, power, and precision we strive for in our database.
Given the logo (and internal company culture around the tiger mascot), I understand where they're coming from, but with the name conflicts (TigerBeetle, WiredTiger, etc) I do wish they'd chosen something else -- like maybe TiScaleDB and give a titanium sheen, do triple duty with the tiger and the Timescale heritage?
You'd be surprised. Tigris is the latinized version of the name of the river in ancient Greek (τίγρης) which also means tiger in ancient Greek.
The common underlying etymology is an even older into-european term translated roughly as "sharp" or "pointy" (in the case of the tiger I guess referring to the teeth).
From a biblical etymology page:
The name Tigris shares its root with the word "tiger" (more precise: the word "tiger" and the name Tigris are identical in Greek). That means that in deep antiquity the tiger and the Tigris had signature qualities that were comparable and from which both derived their name. The word tiger and the identical name Tigris both come from the Avestan word tighri, which means arrow, or the more general tigra, which means sharp or pointed.
Literally last week I was looking at the logo and was like interesting they didn't go with a name using Tiger, Cheetah, etc. Cool name, though I must say Timescale was really cool name as well.
> When we started 8 years ago, SQL databases were “old fashioned.” NoSQL was the future. Hadoop, MongoDB, Cassandra, InfluxDB – these were the new, exciting NoSQL databases. PostgreSQL was old and boring.
In 2017? I thought the NoSQL hype had subsided by then and everyone was excited about distributed transactions -- Spanner, Cockroach, Fauna, Foundation, etc.
> There are no more “SQL vs. NoSQL” debates. MongoDB, Cassandra, InfluxDB, and other NoSQL databases are seen as technical dead ends. Snowflake and Databricks are acquiring PostgreSQL companies. No one talks about Hadoop. The Lakehouse has won.
That's quite some statement. Boy, would I have loved to live in a world where marketing rhetoric and scientific opinion were easier to distinguish.
It sounds totally illogical comment, all those technologies mentioned have only been growing in the last few years and specialised databases are disrupting old school SQL ones.
Cassandra definitely isn’t dead, anyway. InfluxDB is a competitor to Timescale / TigerData, so that’s just a slam on them. I don’t think about MongoDB, other than of course the canonical video [0].
We've been using TimescaleDB/TigerData for over five years now and it has proven to be a reliable component of our project. We process and store hundreds of data points for a six-digit number of industrial robots and TimescaleDB is what makes that possible. While I can't speak for Timescale Cloud, the managed service for TimescaleDB on Azure has been rock solid.
One annoying thing is that tiered storage is not available on their Azure offering, and also in general it feels like managed service for TimescaleDB is the unloved stepchild of their offering.
But yes, I hope the team continues their amazing work, and I'm looking forward to seeing how the project develops in the future.
@jabiko thanks for the note. Glad our product is working so well for you.
re:Azure we are working on some new things :) . Feel free to drop me a message if you'd like to discuss further (ramon@tigerdata.com).
"Why “Tiger”? The tiger has been our mascot since 2017, symbolizing the speed, power, and precision we strive for in our database. Over time, it’s become a core part of our culture: from weekly “Tiger Time” All Hands and monthly “State of the Tiger” business reviews, to welcoming new teammates as “tiger cubs” to the “jungle.”
Meh, it's just a bit of fun. While I don't lean too much into it myself, it's a good way of finding a company where all the grumpy hermits have self-selected themselves away.
> “While I appreciate PostgreSQL every day, am I the only one who thinks this is a rather bad idea?” – top HackerNews comment on our launch (link)
I know it's popular to bash the HackerNews hivemind, and often it's honestly deserved, but this line is in bad taste. The comment was not only polite and professional, it was also right. They had to introduce a columnar storage format (hypertables) to make it work. That is exactly what the comment and the follow-up cocmment suggest.
That's fair. We referenced that quote because it captured a lot of the skepticism in the early days (and because that comment is public). No hard feelings though!
My experiences with Timescale revealed the need for a full time DBA expert of TSDB to make the db viable for queries exceeding more than the last week of time series data. Tiered reads barely work at all. Do you want a degree in how to use a crippled Postgres offshoot?
Tbf, my experience as a DBRE has been that most places should have a DB expert on staff, especially for Postgres. I’ve not used TigerData / Timescale, but IME there’s far more complexity to reason about and manage than people think.
Generally developers need to be watched so they don't blow up the application performance and so they reuse queries in the correct manner so you optimize things like the query cache and the indexes you have.
Query optimization is one of those places where it can be easy to get orders of magnitude performance increases.
Agreed, though it’s also a monstrous effort to get devs to stop chucking everything into JSON, or my new favorite hell, serializing entire classes and storing them as a BLOB.
It’s my fervent belief that we should revert to specialized roles, with a DB team designing schema and queries based on a team’s needs, who can access them via API only. Slows down velocity? Yes. Faster queries and more efficient use of resources? Yes. Fewer incidents and better referential integrity? Also yes.
Slightly off-topic perhaps. For my use case (both short-term and long-term storage of sensors and metrics of a small Home Assistant instance) it probably doesn't matter, but what could someone recommend? ClickHouse looks kind of neat and it doesn't appear to be difficult to admin.
At the company I work at we manage a lot of historical data with Timescale, but we have also had good results with vanilla PostgreSQL for smaller time-series-data-sets. If you are already comfortable with Postgres this might be worth a look.
If you like standard SQL-y type queries, timescaledv itself is a good option. Influx is another option but it has a steeper learning curve and imo it doesn't pay off
I've been using TimescaleDB for a while as a metrics datastore. It's really proven to be great for aggregating data without a lot of hassle (using continuous aggregates, retention policies, etc).
I recommend it when you don't want/need to have separate sources for account data and your metrics/aggregate data.
In a way I guess this makes sense because they are not just doing time series anymore but on the other hand that is just a very strange name. I'm just thinking about Tiger Beetle and I'm sure they will lose so much in brand awareness because people have heard about timescale db but they have not heard about tiger data and the name just sounds so cheesy.
Whenever I see a news headline mentioning that one of my critical dependencies is undergoing changes that make no technical sense, I get real fear. I feel like Jon Snow facing the army, as if I can see a tidal wave of devops work and code fixes coming my way. I really hope that these infrastructure product teams can be considering and not change something that is working well just for the sake of it. Even if they did a really smooth job, that sense of fear itself is a hurt to the brand - making me feel that this product is a source of fear.
justinmitchel | 5 months ago
Cool!
jamesgresql | 5 months ago
jeffchuber | 5 months ago
andyferris | 5 months ago
xeonmc | 5 months ago
jeffchuber | 5 months ago
jitl | 5 months ago
jeffchuber | 5 months ago
Jarwain | 5 months ago
apgwoz | 5 months ago
kajecounterhack | 5 months ago
apgwoz | 5 months ago
* Tiger Shark * Tiger Beetle * Tiger Data * Tiger Games * Tiger Woods * Tiger Attack * Tiger Snake * Wild Tiger
Only one stands out as not like the others. Tiger is too strong a word. The second word disappears.
sidewndr46 | 5 months ago
pcthrowaway | 5 months ago
alexpadula | 5 months ago
evanelias | 5 months ago
tao_oat | 5 months ago
andrenotgiant | 5 months ago
akulkarni | 5 months ago
Also TigerBeetle is an insect, not a fast cat.
apgwoz | 5 months ago
Fair. But a mascot is not a name. I hope you can see why I bring this up?
> Also TigerBeetle is an insect, not a fast cat.
It is? Damn. I thought a Tiger Beetle was a six foot long cat wearing costume wings and a springs for antennae?
agos | 5 months ago
Nezteb | 5 months ago
apgwoz | 5 months ago
evanelias | 5 months ago
apgwoz | 5 months ago
pmalynin | 5 months ago
nelsonfigueroa | 5 months ago
akulkarni | 5 months ago
28304283409234 | 5 months ago
doctoboggan | 5 months ago
Is influxdb really seen as a dead end?
physicles | 5 months ago
lawn | 5 months ago
akulkarni | 5 months ago
kawsper | 5 months ago
While 3. looks impressive, it seems like most of the interesting features are closed source, so not a 1:1 replacement for version 1.
InfluxDB Edge is open-source, but you need to depend on InfluxDB Community which is free, but closed source, to get things like include functionality like a compactor, which will add capabilities for deletes and re-organizing files to optimize for queries on longer time ranges.
They also need to resurrect all their old 1.* Client libraries for 3.*.
I love InfluxDB, but I’m not hopeful for its future.
pauldix | 5 months ago
The 2.x to 3.x move is, admittedly, much harder. This is because of the language Flux. We haven't been able to bring that over to 3.x in a way that makes it useful. We actually built a bridge for it in our cloud offering, but our experience is that the performance isn't good enough to be acceptable for customers wanting to upgrade. If they want to make the move, adopting SQL or InfluxQL is likely the only path.
We'll continue to develop 3.x and we'll build more migration tooling over time. I think we can build specialized tooling to help Flux users migrate over to 3.x with query translation tools, but there are more features we need to land in 3.x to enable that first.
We're committed to the technology stack (Apache Arrow & DataFusion) and the 3.x line. We have no plans for another major release. I'll be happy if we end up releasing 3.56.2 8 years from now.
blitzar | 5 months ago
suyash | 5 months ago
suyash | 5 months ago
ethagnawl | 5 months ago
I did some work using pgvectorscale and their hosted offering a few months back and the product and the team were a delight to work with. I wish TigerData well.
lukaslalinsky | 5 months ago
NewJazz | 5 months ago
jitl | 5 months ago
coldtea | 5 months ago
Not for AI or other bs "pivots"
blitzar | 5 months ago
jamesgresql | 5 months ago
gangstead | 5 months ago
georgewfraser | 5 months ago
I guess it’s just another columnar dbms after all?
yuretz | 5 months ago
dangoodmanUT | 5 months ago
dengolius | 5 months ago
jascha_eng | 5 months ago
In a way Timescale is just postgres on steroids. Sure if you really know your use-case well, are fine with giving up some postgres nicenes, are willing to learn a new query language and are fine with using and syncing multiple data stores you'll outperform timescale. But I think it is still really cool to see how close you can get with essentially just a better postgres.
akulkarni | 5 months ago
"ClickBench evaluates databases using a single table of clickstream data, representative of workloads like web analytics, BI, and log aggregation. It also favors full-table large scans and large-scale aggregations on denormalized data.
Real-time analytics inside applications is different and needs a new benchmark." [0]
This is why we published RTABench. [1]
We believe that it is more representative of real-time analytical workloads.
[0] https://www.tigerdata.com/blog/benchmarking-databases-for-re...
[1] https://rtabench.com/
freilanzer | 5 months ago
qoega | 5 months ago
viccis | 5 months ago
fellatio | 5 months ago
victorbjorklund | 5 months ago
rattray | 5 months ago
> The majority of workloads on our Cloud product aren’t time-series. Companies are running entire applications on us... So we are now “TigerData.” We offer the fastest PostgreSQL. ... Our cloud offering is “Tiger Cloud.” Our logo stays the same: the tiger, looking forward, focused and fast... Our open source time-series PostgreSQL extension remains TimescaleDB. Our vector extension is still pgvectorscale. Why “Tiger”? The tiger has been our mascot since 2017, symbolizing the speed, power, and precision we strive for in our database.
Given the logo (and internal company culture around the tiger mascot), I understand where they're coming from, but with the name conflicts (TigerBeetle, WiredTiger, etc) I do wish they'd chosen something else -- like maybe TiScaleDB and give a titanium sheen, do triple duty with the tiger and the Timescale heritage?
0xdeafbeef | 5 months ago
HackerThemAll | 5 months ago
8K832d7tNmiQ | 5 months ago
nwhnwh | 5 months ago
aduwah | 5 months ago
NewJazz | 5 months ago
bobosha | 5 months ago
totetsu | 5 months ago
talos_ | 5 months ago
aleksi | 5 months ago
I'm pretty sure Tigris Data is named after the Tigris river (https://en.wikipedia.org/wiki/Tigris), and the name does not mean "tiger".
coldtea | 5 months ago
The common underlying etymology is an even older into-european term translated roughly as "sharp" or "pointy" (in the case of the tiger I guess referring to the teeth).
From a biblical etymology page:
The name Tigris shares its root with the word "tiger" (more precise: the word "tiger" and the name Tigris are identical in Greek). That means that in deep antiquity the tiger and the Tigris had signature qualities that were comparable and from which both derived their name. The word tiger and the identical name Tigris both come from the Avestan word tighri, which means arrow, or the more general tigra, which means sharp or pointed.
alexpadula | 5 months ago
cakoose | 5 months ago
In 2017? I thought the NoSQL hype had subsided by then and everyone was excited about distributed transactions -- Spanner, Cockroach, Fauna, Foundation, etc.
politelemon | 5 months ago
akulkarni | 5 months ago
"The future is already here, it's just not very evenly distributed" - William Gibson
dgellow | 5 months ago
spooneybarger | 5 months ago
smokel | 5 months ago
That's quite some statement. Boy, would I have loved to live in a world where marketing rhetoric and scientific opinion were easier to distinguish.
baggiponte | 5 months ago
suyash | 5 months ago
mellosouls | 5 months ago
There's an element of immaturity in the style that they should probably work on.
inamorty | 5 months ago
sgarland | 5 months ago
[0]: https://youtu.be/b2F-DItXtZs
jabiko | 5 months ago
One annoying thing is that tiered storage is not available on their Azure offering, and also in general it feels like managed service for TimescaleDB is the unloved stepchild of their offering.
But yes, I hope the team continues their amazing work, and I'm looking forward to seeing how the project develops in the future.
ramonguiu | 5 months ago
morelish | 5 months ago
conradev | 5 months ago
igitur | 5 months ago
ovaistariq | 5 months ago
gagik_co | 5 months ago
rbaudibert | 5 months ago
LeonM | 5 months ago
Care to elaborate on why you posted this?
aorth | 5 months ago
Did you try Slack?
v5v3 | 5 months ago
Cringe...!!!
lijok | 5 months ago
Freak_NL | 5 months ago
1: https://media.karousell.com/media/photos/products/2018/11/10...
coldtea | 5 months ago
coldtea | 5 months ago
matsemann | 5 months ago
freilanzer | 5 months ago
heeton | 5 months ago
dzonga | 5 months ago
akulkarni | 5 months ago
eska | 5 months ago
I know it's popular to bash the HackerNews hivemind, and often it's honestly deserved, but this line is in bad taste. The comment was not only polite and professional, it was also right. They had to introduce a columnar storage format (hypertables) to make it work. That is exactly what the comment and the follow-up cocmment suggest.
akulkarni | 5 months ago
eska | 5 months ago
zlib | 5 months ago
koakuma-chan | 5 months ago
halfmatthalfcat | 5 months ago
Dowwie | 5 months ago
akulkarni | 5 months ago
sgarland | 5 months ago
pixl97 | 5 months ago
Query optimization is one of those places where it can be easy to get orders of magnitude performance increases.
sgarland | 5 months ago
It’s my fervent belief that we should revert to specialized roles, with a DB team designing schema and queries based on a team’s needs, who can access them via API only. Slows down velocity? Yes. Faster queries and more efficient use of resources? Yes. Fewer incidents and better referential integrity? Also yes.
orphea | 5 months ago
ejs | 5 months ago
skowalak | 5 months ago
suyash | 5 months ago
suyash | 5 months ago
Rebelgecko | 5 months ago
ejs | 5 months ago
I recommend it when you don't want/need to have separate sources for account data and your metrics/aggregate data.
victorbjorklund | 5 months ago
tianqi | 5 months ago
geodel | 5 months ago
ahmadtbk | 5 months ago
https://chatgpt.com/share/6852de93-1384-8004-ac63-4ae93a8373...