Author here: Adding some context. I led the Postgres platform team (2019-23) at Cloudflare and we were supporting 170+ growing product teams. One of the constant asks is schema migration review. We published a lot of best practices, added CI checks however, it was still hard to catch. Also, I tried to explain the internals of how the locking (rewrite) works, but I realized most of the devs just want the answer - Is it safe or not safe to run?
Not sure if it rings a bell, the name is a reference to the Silicon Valley Jian Yang's hot dog or not hot dog app.
Also, I understand the decision of safe vs not-safe depends heavily on data/histogram and edge cases, but still quite a lot of low-hanging issues can be easily caught with a deterministic rule engine. So I ported pg_savior[1] and used sql parser from libpg-query-node[2] which compiles as WASM, so it entirely runs on the browser. No telemetry, no login. Source attached [3]
It's not immediately clear from the README, but is it easy to run with multiple profiles like "backwards-compatible", "revertable" (both data and schema) and "destructive" for that final clean-up in multi-staged no-downtime migrations? Basically common subsets of "safe-ness" of the schema migration queries.
I imagine it can be tuned, but I'd love this for all my projects.
And since I am currently on a project doing MS SQL (gasp), that'd be cool too ;)
I am familiar with an "is it a hot dog" app from back in the day, bit would have never made the connection :)
Thing that bit me most wasn't the DDL itself, it was lock queuing. An ADD COLUMN is instant but if it waits behind a long read, every query behind it piles up too. Lock_timeout plus retry saved us more than any clever migration tool.
Author here: Adding some context. I led the Postgres platform team (2019-23) at Cloudflare and we were supporting 170+ growing product teams. One of the constant asks is schema migration review. We published a lot of best practices, added CI checks however, it was still hard to catch. Also, I tried to explain the internals of how the locking (rewrite) works, but I realized most of the devs just want the answer - Is it safe or not safe to run?
Not sure if it rings a bell, the name is a reference to the Silicon Valley Jian Yang's hot dog or not hot dog app.
Also, I understand the decision of safe vs not-safe depends heavily on data/histogram and edge cases, but still quite a lot of low-hanging issues can be easily caught with a deterministic rule engine. So I ported pg_savior[1] and used sql parser from libpg-query-node[2] which compiles as WASM, so it entirely runs on the browser. No telemetry, no login. Source attached [3]
[1] https://github.com/viggy28/pg_savior [2] https://github.com/constructive-io/libpg-query-node [3] https://github.com/viggy28/safe-not-safe
Wow, great idea!
It's not immediately clear from the README, but is it easy to run with multiple profiles like "backwards-compatible", "revertable" (both data and schema) and "destructive" for that final clean-up in multi-staged no-downtime migrations? Basically common subsets of "safe-ness" of the schema migration queries.
I imagine it can be tuned, but I'd love this for all my projects.
And since I am currently on a project doing MS SQL (gasp), that'd be cool too ;)
I am familiar with an "is it a hot dog" app from back in the day, bit would have never made the connection :)
Very cool! I wrote a go library for this https://onwardpg.solberg.is
Nice. Integrating it on CI and catching there is still the best way/place. Having said, it doesn't work like that in practice.
really like the browser-only approach here - catching the obvious migration risks locally before they ever reach CI feels super useful
Thing that bit me most wasn't the DDL itself, it was lock queuing. An ADD COLUMN is instant but if it waits behind a long read, every query behind it piles up too. Lock_timeout plus retry saved us more than any clever migration tool.
Agree. `lock_timeout` will go a long way in terms of damage control.