Any FoundationDB event planned?

Community
1

Is there a community event that is planned in the coming months? What about FoundationDB Summit?

2

I am also wondering about the same.

It would also be nice if we could have regular monthly or quarterly meetups (online or offline) where we can share some of the things that we are working on. :slight_smile:

3

:+1:

I do think the FDB’s community needs to have some regular online meetups. @amirouche actually started the FoundationDB Paris Meetup, and I remember having to switch to English because people who do not speak French had joined.

I think this is a separate question, compared to the possibility to meet (online) once for a while :smile: FoundationDB Summit 2018 was my entrypoint to FDB, and I would love to see it coming back.

4

I am not comfortable doing a meetup without a talk.

@rajivr do you want to do a presentation in an online event over Google Meet at 10:00 AM PDT? It will be 7 PM in CET that would be perfect for a meetup. If yes, what day in May will be best for you?

5

Sure! I’ll be happy to give a talk about my work. Let me DM you.

3 Likes
6

Thanks @rajivr

7

The slides from the meetup is here.

2 Likes
8

I lost the screen record. I invite you to record again yourself (improving the transcript with the discussion that happened); I did one in french at : hyper.dev (I proof read the google translation)

My take-aways from the whole event:

  1. Check recording machinery twice, one or two days before the event;

  2. Transactions with timeouts and retries can write the same transaction twice if they are not idempotent;

  3. VersionStamps go into tuples, they are promised for a unique value based on the transaction number (mind the fact that it may break when restoring a backup, and importing data from another cluster);

  4. There is still accidental complexity e.g. 2) transactions with timeouts, or [1]; but on average FoundationDB solve most of the essential complexity when dealing with to scale data.

[1] In python bindings, the fact that None is mocked up, leads to the code if Key == None which is not green horn proof, because None is singleton, and the comparison with singleton is done like if object is singleton. That at its time, yields to various questions about how it python mocks are done, and what is duck typing. That conversation might be necessary even if absent keys were the None.