HNHacker News
TopNewBestAskShowJobs

gauravphoenix

591 karma · joined June 5, 2014

Security dude. www.linkedin.com/in/gauravphoenix
submissionscomments
gauravphoenix··on Bank Run in Los Angeles at a First Republic Bank Branch
FYI: https://twitter.com/firstrepublic/status/1635005464584024064...

The bank is plenty liquid and has unused $60B credit line with fed. I hope we can use social media to educate people and not partake in creating another bank run.

gauravphoenix··on Bank Run in Los Angeles at a First Republic Bank Branch
Why would you move $ out regardless then? The very fact that you are moving $ out after reading on social media means that you are giving in to the panic being created by people who have shorted stocks of FRC etc.
gauravphoenix··on 30% of YC companies exposed through SVB can’t make payroll in the next 30 days
As of right now, yes. Give it a week, and it will get sorted out. In all likelihood, the bank will have a new owner by coming Monday/Tuesday and everyone will be able to make the transactions.
gauravphoenix··on Office Landlord Defaults Are Escalating as Lenders Brace for More Distress
I am an investor via CrowdStreet in one of the office buildings in south-ish area (Campbell) of silicon valley. Here is a market update I got today-

  Silicon Valley’s office vacancy rate increased again in Q4 2022 to 19.0%, up from 18.6% in Q3. The end of 2022 marks the 12th consecutive   
  quarter where the overall vacancy rate has risen. Campbell’s vacancy rate for in Q4 was 30.0% with negative 278,167 of overall absorption. 
  Silicon Valley’s average asking rent increased slightly in Q4 2022, rising to $5.36 psf on a monthly full-service basis, a $0.02 increase   
  from the previous quarter and a $0.04 increase year-over-year. Campbell’s average asking rent was $4.18. Given the current lack of demand 
  for office space as well as upcoming economic uncertainty, there is little expectation that significant rent increases would occur in the  
  near term. Transaction volume fell sharply in Q4 2022, posting just 778,712 sf of gross absorption, a 76% decrease compared to the previous 
  quarter. Demand for new office space fell on the news of large-scale tech-sector layoffs this quarter that will likely persist through 2023. 
  The Valley measured 2.6 msf of net occupancy losses in 2022, the largest amount since 2001 and marks the third year of negative net 
  absorption
gauravphoenix··on Quitting the rat race
Opinions are never right or wrong. That's by definition.

Facts on the other hand can be right or wrong as they can be independently verified.

There is no way to prove "great" things cities can provide. Different people value different things.

gauravphoenix··on Quitting the rat race
>Hey, if being in nature makes you happy good for you but your outlook on the rest is kind of crap.

It reads like you are stating a fact, but really, it is a matter of opinion. Perhaps, you can just say what you like without invalidating what others prefer.

gauravphoenix··on Show HN: Publish from GitHub Actions using multi-factor authentication
Congrats Varun :) Great to see my ex-roommate's post on HN front page.

Fun fact- Varun is a super smart engineer, he even won a car in a competition:

https://www.oneindia.com/2006/06/29/microsoft-security-shoot...

I would highly encourage HNers to try out his github action.

gauravphoenix··on “Invalid Username or Password”: a useless security measure (2014)
>spam users with nonsense like product updates

IMHO, not all product updates are nonsense.

gauravphoenix··on WhatsApp Remote Code Execution in Video Call
What is the worst case scenario here? Will the adversary be able to break out of the sandbox? i.e. will the adversary be able to access non WhatsApp data?
gauravphoenix··on Search 5.8B images used to train popular AI art models
interestingly, if you search for "127.0.0.1", it throws

  {"message":"Forbidden"}
in the api response and the page says-

  Sorry, there was an error with your search. Please try a different request.
gauravphoenix··on Crazy Thin ‘Deep Insert’ ATM Skimmers
Not a payment security expect, but based on my limited knowledge, the transaction is verified with the chip embedded in the card/device itself so I don't think that skimming devices like these can bypass NFC protections.
gauravphoenix··on ‘How to Tell a Story’ Review: Aristotle in the Writers’ Room
link to substack: https://georgesaunders.substack.com
gauravphoenix··on No, QuestDB is not Faster than ClickHouse
I have always felt that DB benchmarks are useless, always use your own dataset

https://gauravkumar.blog/performance-benchmarks-are-useless....

gauravphoenix··on Lesser known features of ClickHouse
> And your materialised view/JsonExtract can be dispatched by schema name for a row.

How would that look like? MatViews in CH are like insert triggers so does it mean that one MatView per schema?

>I see the only suboptimal part if different schemas have different types for the same field and it has to convert it to String.

Agree, this is what I was actually after and it actually happens with system data sometimes- you can have "user" field as an int and string in different schemas.

gauravphoenix··on Lesser known features of ClickHouse
>3. Hosting static MergeTree tables on any web server.

Can you share some docs on that? AFAIK CH can use web server hosted data sources but I don't recall CH able to host merge tree tables on HTTP servers.

gauravphoenix··on Lesser known features of ClickHouse
CH is actually very good if you can model the dimensional data as dictionary. Also CH can use external store like PSQL as dictionary. But yes, if your fact table data size is as big as dimensional data, CH or any other such system will be quite slow.
gauravphoenix··on Lesser known features of ClickHouse
If you use the "Object" data type (json), you can have a schema-less datastore and that can alleviate the need of complex ETL pipelines. But yes, if you need to decorate the data before ingestion, you need some ETL pipeline. In other words, if the ETL pipeline is just storing the data in columnar format by reading some JSON fields, you don't need that anymore with the new CH json data type
gauravphoenix··on Lesser known features of ClickHouse
the "Object" data type works very well and is quite performant. But do note that if you have lots of "schemas" you will need to create multiple such columns that represent such columns e.g. one column each for GitHub,GitLab etc.
gauravphoenix··on Lesser known features of ClickHouse
We at Dassana[1] use clickhouse underneath and solve the exact same problem. Have a look and feel free to reach out to me if you have any questions. You can find more details on my ShowHN post[2].

[1] https://lake.dassana.io

[2] https://news.ycombinator.com/item?id=31111432

gauravphoenix··on Oral administration of NMN is safe and efficiently increases blood NAD levels
Do note conflict of interest:

>YU, YF, TaS, and ToS were employees of Mitsubishi Corporation Life Sciences Limited and prepared NMN and the placebo

gauravphoenix··on Zq: An easier and faster alternative to jq
there is Dassana[1] if someone wants to try out json native,index-free, schema-less solution built on top of ClickHouse.

ShowHN post(FAQ)[2]

disclaimer- I'm founder/CEO of Dassana.

[1] https://lake.dassana.io/

[2] https://news.ycombinator.com/item?id=31111432

gauravphoenix··on Show HN: Dassana. JSON-native,schema-less logging solution built atop ClickHouse
we flatten nested JSONs too. Not sure what you meant by second question, can you rephrase or provide an example? Also, feel free to drop me an email or join slack[1]- it will be easier to discuss tech details there.

[1] https://dassanacommunity.slack.com/join/shared_invite/zt-teo...

gauravphoenix··on Show HN: Dassana. JSON-native,schema-less logging solution built atop ClickHouse
We flatten it to a map and store the map. When you query, we dynamically generate ClickHouse query that queries the map.
gauravphoenix··on Show HN: Dassana. JSON-native,schema-less logging solution built atop ClickHouse
A few differences:

- Even though logtail is quite cheap, we are even cheaper solution - Our pricing model separates ingestion from query. If you don't query, you don't pay for query. Just ingestion. - We are JSON native. Our SQL allows querying JSON fields that are nested under nested JSON arrays. - Performance. We believe our solution is much faster for selective queries. This like most performance claims, it all just depends on the data shape, volume and what you are querying.

Similarities: - We both use ClickHouse as underlying DB.

gauravphoenix··on Show HN: Dassana. JSON-native,schema-less logging solution built atop ClickHouse
Generally speaking, tech like ClickHouse is considered a data warehouse tech. But we are using it as data lake tech: there are no schemas involved in Dassana. This means that you can send free form JSON objects to Dassana and query them using SQL.
gauravphoenix··on Show HN: Dassana. JSON-native,schema-less logging solution built atop ClickHouse
Depends- there is plenty of market for purely structured logs query. I come from security background and these days in cloud security, all logs are structured: CloudTrail,VpcFlows,ALB logs etc. We are going to be focusing on such use cases.
gauravphoenix··on Show HN: Dassana. JSON-native,schema-less logging solution built atop ClickHouse
>Who do you see as your competition? AWS's CloudWatch / Centralized Logging? Splunk? GCP's Logging? Logstash? Graylog?

More like Athena/Presto/SnowFlake. Simply put, anyone who is DB like systems for querying structured logs.

>What kind of query language are you thinking? I imagine SQL-like, as that's Clickhouse's native language.

Pretty much CH like SQL with some syntactic sugar for JSON. https://docs.dassana.cloud/docs/query/sample-queries#filter-...

>Business-wise, how are you gonna integrate with the cloud providers, AWS / GCP / Azure? Most people who use those services just use the built-ins.

Cheaper, faster and easier. Let me openly challenge anyone- take a nested JSON, send it to logging service and query it. And now do the same with Dassana. You will find night and day difference. I agree with most folks start with in-built services but they soon grow out of it. Here is an example- if you are using GCP, try getting count of failed HTTP requests group by host or ip. Turns out, there is no support for aggregate queries and you will have to create bunch of complicated metric filters to achieve it.

>More than Grafana, I think you need something like Metabase integrated OOTB. That might be a killer feature.

That's an awesome suggestion, we are going to look into it

>IMHO, FTS is a must-have from day 1. Most software that folks run produce non-structured logs OOTB (sad, I know), so folks won't even be able to try your service without changing their software. And getting a lot of software, even popular ones like Python/Flask, Ruby/Rails, Java/Spring, to produce structured logs is not a simple task.

Agree with your sentiment,we for now are focusing on use cases where you have structured logs. SecOps teams have such use cases. These teams mostly deal with data like CloutTrail, VpcFlows, ALB logs etc.

gauravphoenix··on Show HN: Dassana. JSON-native,schema-less logging solution built atop ClickHouse
Not yet. It is quite a bit rough around the edges and far from production use. Besides, it will require creation of multiple tables/column for each schema type (i.e. github schema might conflict with gitlab schema). As such, we decided to flatten the JSONs and use map data type and have our own SQL layer that translates the queries to underlying ClickHouse queries. This allows us to add a lot of syntactic sugar e.g. https://docs.dassana.cloud/docs/query/intro

We might start using that feature in future.

gauravphoenix··on SDF Public Access Unix System
oh the memories! I used sdf more than two decades ago and learnt so much about internet from it.

On a side note- interesting to see the copyright: ©1987-2065

gauravphoenix··on IPv6-ghost-ship: Using TOTP as part of an AWS EC2 IPv6 address
Aidan is one of the most brilliant engineering person I have known. I share a slack workspace with him and his posts are always enlightening.

Glad to see him on HN front-page.

← PreviousPage 2 of 5Next →