Hi All!
Thank you for the answers and concerns from everyone.
On the CLI vs State Metadata Connector/Table question I would also like to
step back a little and look at the bigger picture.
I think the overall vision in Flink SQL is to provide a SQL native
environment where we can serve complex use-cases like you would expect in a
regular database.
Most features, developments in the recent years have gone this way.
The State Metadata Table would be a natural and straightforward fit here.
So from my side, +1 for that.
However I could understand if we are not ready to add a new
connector/format due to maintenance concerns (and in general concern about
the design).
If that's the issue then we should spend more time on the design to get
comfortable with the approach and seek feedback from the wider community
I am -1 for the CLI/tooling approach as that will not provide the
featureset we are looking for that is not already covered by the Java
connector. And that approach would come with the same maintenance
implications.
Cheers
Gyula
On Wed, Mar 19, 2025 at 11:24 AM Gabor Somogyi <gabor.g.somo...@gmail.com>
wrote:
> Hi Zaklely, Shengkai
>
> Several topics are going on so adding gist answers to them. When some topic
> is not touched please highlight it.
>
> @Shengkai: I've read through all the previous FLIPs related catalogs and if
> we would like to keep the concepts there
> then one-to-one mapping relationship between savepoint and catalog is a
> reasonable direction. In short I'm happy that
> you've highlighted this and agree as a whole. I've written it down
> previously, just want to double confirm that state catalog is
> essential and planned. When we reach this point then your input is more
> than welcome.
>
> @Zakelly: We've tried the CLI and separate library approaches with users
> already and these are not something which is welcome because of the
> following:
> * Users want to have automated tasks and not manual CLI/library output
> parsing. This can be hacked around but our experience is negative on this
> because it's just brittle.
> * From development perspective It's way much bigger effort than a connector
> (hard to test, packaging/version handling is and extra layer of complexity,
> external FS authentication is pain for users, expecting them to download
> savepoints also)
> * Purely personal opinion but if we would find better ways later then
> retire a CLI is not more lightweight than retire a connector
>
> > It would be great if you give some examples on how user could leverage
> the separate connector to process the metadata.
>
> The most simplest cases:
> * give me the overgroving state uids
> * give me the not known (new or renamed) state uids
> * give me the state uids where state size drastically dropped compare to a
> previous savepoint (accidental state loss)
>
> Since it was mentioned: as a general offtopic teaser, yeah it would be good
> to have some sort of checkpoint/savepoint lineage or however we call it.
> Since we've not yet reached this point there are no technical details, it's
> more like a vision. It's a common pattern that
> jobs are physically running but somehow the state processing is stuck and
> it would be good to add some way to find it out automatically.
> The important saying here is automation and not manual evaluation since
> handling 10k+ jobs is just not allowing that.
>
> BR,
> G
>
>
> On Wed, Mar 19, 2025 at 6:46 AM Shengkai Fang <fskm...@gmail.com> wrote:
>
> > Hi, All.
> >
> > About State Catalog, I want to share more thoughts about this.
> >
> > In the initial design concept, I understood that a savepoint and a state
> > catalog have a one-to-one mapping relationship. Each operator corresponds
> > to a database, and the state of each operator is represented as
> individual
> > tables. The rationale behind this design is:
> >
> > *State Diversity*: An operator may involve multiple types of states. For
> > example, in our VVR design, a "multi-join" operator uses keyed states for
> > two input streams and a broadcast state for the third stream. This makes
> it
> > challenging to represent all states of an operator within a single table.
> > *Scalability*: Internally, an operator might have multiple keyed states
> > (e.g., value state and list state). However, large list states may not
> fit
> > entirely in memory. To address this, we recommend implementing each state
> > as a separate table.
> >
> > To resolve the loosely coupled relationships between operator states, we
> > propose embedding predefined views within the catalog. These views
> simplify
> > user understanding of operator implementations and provide a more
> intuitive
> > perspective. For instance, a join operator may have multiple state
> > implementations (depending on whether the join key includes unique
> > attributes), but users primarily care about the data associated with a
> > specific join key across input streams.
> >
> > Returning to the one-to-one mapping between savepoints and catalogs, we
> aim
> > to manage multiple user state catalogs through a catalog store. When a
> user
> > triggers a savepoint for a job on the platform:
> >
> > 1. The platform sends a REST request to the JobManager.
> > 2. Simultaneously, it registers a new state catalog in the catalog store,
> > enabling immediate analysis of state data on the platform.
> > 3. Deleting a savepoint would also trigger the removal of its associated
> > catalog.
> >
> > This vision assumes that states are self-describing or that a state
> > metaservice is introduced to analyze savepoint structures.
> >
> > > How can users create logic to identify differences between multiple
> > savepoints?
> >
> > Since savepoints and state catalogs are one-to-one mapped, users can
> query
> > metadata via their respective catalogs. For example:
> >
> > 1. `savepoint-${id}`.`system`.`metadata_table`.`<operator-name>` provides
> > operator-specific metadata (e.g., state size, type).
> > 2. Comparing metadata tables (e.g., schema versions, state entry counts)
> > across catalogs reveals structural or quantitative differences.
> > 3. For deeper analysis, users could write SQL queries to compare specific
> > state partitions or leverage the metaservice to track state evolution
> > (e.g., added/removed operators, modified state configurations).
> >
> > If we plan to introduce a state catalog in the future, I would lean
> toward
> > using metadata tables. If a utility tool can address the challenges we
> > face, could we avoid introducing an additional connector?
> >
> > Best,
> > Shengkai
> >
> > Gyula Fóra <gyula.f...@gmail.com> 于2025年3月17日周一 20:25写道：
> >
> > > Hi All!
> > >
> > > Without going into too much detail here are my 2 cents regarding the
> > > virtual column / catalog metadata / table (connector) discussion for
> the
> > > State metadata.
> > >
> > > State metadata such as the types of states, their properties, names,
> > sizes
> > > etc are all valuable information that can be used to enrich the
> > > computations we do on state.
> > > We can either analyze it standalone (such as discover anomalies, for
> > large
> > > jobs with many states), across multiple savepoints (discover how state
> > > changed over time) or by joining it with keyed or non-keyed state data
> to
> > > serve more complex queries on the state.
> > >
> > > The only solution that seems to serve all these use-cases and
> > requirements
> > > in a straightforward and SQL canonical way is to simply expose the
> state
> > > metadata as a separate table. This is a metadata table but you can also
> > > think of it as data table, it makes no practical difference here.
> > >
> > > Once we have a catalog later, the catalog can offer this table out of
> the
> > > box, the same way databases provide metadata tables. For this to work
> > > however we need another, simpler connector that creates this table.
> > >
> > > +1 for state metadata as a separate connector/table, instead of adding
> > > virtual columns and adhoc catalog metadata that is hard to use in a
> large
> > > number of queries.
> > >
> > > Cheers,
> > > Gyula
> > >
> > > On Mon, Mar 17, 2025 at 12:44 PM Gabor Somogyi <
> > gabor.g.somo...@gmail.com>
> > > wrote:
> > >
> > > > 1. State TTL for Value Columns
> > > >
> > > > > I’m planning on adding this, and we may collaborate on it in the
> > > future.
> > > >
> > > > +1 on this, just ping me.
> > > >
> > > > 2. Metadata Table vs. Metadata Column
> > > >
> > > > After some code digging and POC all I can say that with heavy effort
> we
> > > can
> > > > maybe add such changes that we're able to show metadata of a
> savepoint
> > > from
> > > > catalog.
> > > > I'm not against that but from user perspective this has limited
> value,
> > > let
> > > > me explain why.
> > > >
> > > > From high level perspective I see the following which I see agreement
> > on:
> > > > * We should have a catalog which is representing one or more jobs
> > > savepoint
> > > > data set (future plan)
> > > > * Savepoints should be able to be registered in the catalog which are
> > > then
> > > > databases (future plan)
> > > > * There must be a possiblity to create tables from databases where
> > users
> > > > can read state data (exists already)
> > > >
> > > > In terms of metadata, If I understand correctly then the suggested
> > > approach
> > > > would be to access
> > > > it from the catalog describe command, right? Adding that info when
> > > specific
> > > > database describe command
> > > > is executed could be done.
> > > >
> > > > The question is for instance how can users create such a logic that
> > tells
> > > > them what is
> > > > the difference between multiple savepoints?
> > > > Just to give some examples:
> > > > * per operator size changes between savepoints
> > > > * show values from operator data where state size reaches a boundary
> > > > * in general "find which checkpoint ruined things" is quite common
> > > pattern
> > > > What I would like to highlight here is that from Flink point of view
> > the
> > > > metadata can be
> > > > considered as a static side output information but for users these
> > values
> > > > are actual real data
> > > > where logic is planned to build around.
> > > >
> > > > > The metadata is more like one-time information instead of a
> streaming
> > > > data that changes all
> > > > the time, so a single connector seems to be an overkill.
> > > >
> > > > State data is also static within a savepoint and that's the reason
> why
> > > the
> > > > state processor API is working in batch mode.
> > > > When we handle multiple checkpoints in a streaming fashion then this
> > can
> > > be
> > > > viewed from another angle.
> > > >
> > > > We can come up with more lightweight solution other than a new
> > connector
> > > > but enforcing users to parse the catalog
> > > > describe command output in order to compare multiple savepoints
> doesn't
> > > > sound smooth user experience.
> > > > Honestly I've no other idea how exposing metadata as real user data
> so
> > > > waiting on other approaches.
> > > >
> > > > BR,
> > > > G
> > > >
> > > >
> > > > On Thu, Mar 13, 2025 at 2:44 AM Shengkai Fang <fskm...@gmail.com>
> > wrote:
> > > >
> > > > > Looking forward to hearing the good news!
> > > > >
> > > > > Best,
> > > > > Shengkai
> > > > >
> > > > > Gabor Somogyi <gabor.g.somo...@gmail.com> 于2025年3月12日周三 22:24写道：
> > > > >
> > > > > > Thanks for both the valuable input!
> > > > > >
> > > > > > Let me take a closer look at the suggestions, like the Catalog
> > > > > capabilities
> > > > > > and possibility of embedding TypeInformation or
> > > > > > StateDescriptor metadata directly into the raw state files...
> > > > > >
> > > > > > BR,
> > > > > > G
> > > > > >
> > > > > >
> > > > > > On Wed, Mar 12, 2025 at 8:17 AM Shengkai Fang <fskm...@gmail.com
> >
> > > > wrote:
> > > > > >
> > > > > > > Thanks for Zakelly's clarification.
> > > > > > >
> > > > > > > 1. State TTL for Value Columns
> > > > > > >
> > > > > > > +1 to delay the discussion about this.
> > > > > > >
> > > > > > > 2. Metadata Table vs. Metadata Column
> > > > > > >
> > > > > > > I’d like to share my perspective on the State Catalog proposal.
> > > While
> > > > > > > introducing this capability is beneficial, there is a blocker:
> > the
> > > > > > current
> > > > > > > StateBackend architecture does not permit operators to encode
> > > > > > > TypeInformation into the state—it only preserves the
> Serializer.
> > > This
> > > > > > > limitation creates an asymmetry, as operators alone retain
> > > knowledge
> > > > of
> > > > > > the
> > > > > > > data structure’s schema.
> > > > > > >
> > > > > > > To address this, I suggest allowing operators to embed
> > > > TypeInformation
> > > > > or
> > > > > > > StateDescriptor metadata directly into the raw state files.
> Such
> > a
> > > > > design
> > > > > > > would enable the Catalog to:
> > > > > > >
> > > > > > > 1. Parse state files and programmatically derive the schema and
> > > > > > structural
> > > > > > > guarantees for each state.
> > > > > > > 2. Leverage existing Flink Table utilities, such as
> > > > > > > LegacyTypeInfoDataTypeConverter (in
> > > > > org.apache.flink.table.types.utils),
> > > > > > to
> > > > > > > bridge TypeInformation and DataType conversions.
> > > > > > >
> > > > > > > If we can not store the TypeInformation or StateDescriptor into
> > the
> > > > raw
> > > > > > > state files, I am +1 for this FLIP to use metadata column to
> > > retrieve
> > > > > > > information.
> > > > > > >
> > > > > > > Best,
> > > > > > > Shengkai
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > Zakelly Lan <zakelly....@gmail.com> 于2025年3月12日周三 12:43写道：
> > > > > > >
> > > > > > > > Hi Gabor and Shengkai,
> > > > > > > >
> > > > > > > > Thanks for sharing your thoughts! This is a long discussion
> and
> > > > sorry
> > > > > > for
> > > > > > > > the late reply (I'm busy catching up with release 2.0 these
> > > days).
> > > > > > > >
> > > > > > > > 1. State TTL for Value Columns
> > > > > > > >
> > > > > > > >
> > > > > > > > Let me first clarify your thoughts to ensure I understand
> > > > correctly.
> > > > > > > IIUC,
> > > > > > > > there is no persistent configuration for state TTL in the
> > > > checkpoint.
> > > > > > > While
> > > > > > > > you can infer that TTL is enabled by reading the serializer,
> > the
> > > > > > > checkpoint
> > > > > > > > itself only stores the last access time for each value. So
> the
> > > only
> > > > > > thing
> > > > > > > > we can show is the last access time for each value. But it is
> > not
> > > > > > > required
> > > > > > > > for all state backends to store this, as they may directly
> > store
> > > > the
> > > > > > > > expired time. This will also increase the difficulty of
> > > > > implementation
> > > > > > &
> > > > > > > > maintenance.
> > > > > > > >
> > > > > > > > This once again reiterates the importance of unified metadata
> > for
> > > > > > > > checkpoints. I’m planning on adding this, and we may
> > collaborate
> > > on
> > > > > it
> > > > > > in
> > > > > > > > the future.
> > > > > > > >
> > > > > > > > 2. Metadata Table vs. Metadata Column
> > > > > > > >
> > > > > > > >
> > > > > > > > I'm not in favor of adding a new connector for metadata. The
> > > > metadata
> > > > > > is
> > > > > > > > more like one-time information instead of a streaming data
> that
> > > > > changes
> > > > > > > all
> > > > > > > > the time, so a single connector seems to be an overkill. It
> is
> > > not
> > > > > easy
> > > > > > > to
> > > > > > > > withdraw a connector if we have a better solution in future.
> > I'm
> > > > not
> > > > > > > > familiar with current Catalog capabilities, and if it could
> > > extract
> > > > > and
> > > > > > > > show some operator-level information from savepoint, that
> would
> > > be
> > > > > > great.
> > > > > > > >
> > > > > > > > If the Catalog can't do that, I would consider the current
> FLIP
> > > to
> > > > > be a
> > > > > > > > compromise solution.
> > > > > > > >
> > > > > > > > And if we have that unified metadata for checkpoint/savepoint
> > in
> > > > > > future,
> > > > > > > we
> > > > > > > > may directly register savepoint in catalog, and create a
> source
> > > > > without
> > > > > > > > specifying complex columns, as well as describe the savepoint
> > > > catalog
> > > > > > to
> > > > > > > > get the metadata. That's a good solution in my mind.
> > > > > > > >
> > > > > > > >
> > > > > > > > Best,
> > > > > > > > Zakelly
> > > > > > > >
> > > > > > > > On Wed, Mar 12, 2025 at 10:35 AM Shengkai Fang <
> > > fskm...@gmail.com>
> > > > > > > wrote:
> > > > > > > >
> > > > > > > > > Hi Gabor,
> > > > > > > > >
> > > > > > > > > > 2. Adding a new connector with `savepoint-metadata`
> > > > > > > > >
> > > > > > > > > I would argue against introducing a new connector type
> named
> > > > > > > > > savepoint-metadata, as the existing Catalog mechanism can
> > > > > inherently
> > > > > > > > > provide the necessary connector factory capabilities. I’ve
> > > > detailed
> > > > > > > this
> > > > > > > > > proposal in branch[1]. Please take a moment to review it.
> > > > > > > > >
> > > > > > > > > If we introduce a connector named `savepoint-metadata`, it
> > > means
> > > > > user
> > > > > > > can
> > > > > > > > > create a temporary table with connector
> `savepoint-metadata`
> > > and
> > > > > the
> > > > > > > > > connector needs to check whether table schema is same to
> the
> > > > schema
> > > > > > we
> > > > > > > > > proposed in the FLIP. On the other hand, it's not easy work
> > for
> > > > > > others
> > > > > > > to
> > > > > > > > > users a metadata table with same schema.
> > > > > > > > >
> > > > > > > > > [1]
> > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
> https://github.com/apache/flink/compare/master...fsk119:flink:state-metadata?expand=1#diff-712a7bc92fe46c405fb0e61b475bb2a005cb7a72bab7df28bbb92744bcb5f465R63
> > > > > > > > >
> > > > > > > > > Best,
> > > > > > > > > Shengkai
> > > > > > > > >
> > > > > > > > > Gabor Somogyi <gabor.g.somo...@gmail.com> 于2025年3月11日周二
> > > 16:56写道：
> > > > > > > > >
> > > > > > > > > > Hi Shengkai,
> > > > > > > > > >
> > > > > > > > > > > 1. State TTL for Value Columns
> > > > > > > > > >
> > > > > > > > > > From directional perspective I agree your idea how it can
> > be
> > > > > > > > implemented.
> > > > > > > > > > Previously I've mentioned that TTL information is not
> > exposed
> > > > on
> > > > > > the
> > > > > > > > > state
> > > > > > > > > > processor API (which the SQL state connector uses to read
> > > data)
> > > > > > > > > > and unless somebody show me the opposite this FLIP is not
> > > going
> > > > > to
> > > > > > > > > address
> > > > > > > > > > this to avoid feature creep. Our users are also
> interested
> > in
> > > > TTL
> > > > > > so
> > > > > > > > > > sooner or later we're going to expose it, this is matter
> of
> > > > > > > scheduling.
> > > > > > > > > >
> > > > > > > > > > > 2. Adding a new connector with `savepoint-metadata`
> > > > > > > > > >
> > > > > > > > > > Not sure I understand your point at all related
> > StateCatalog.
> > > > > First
> > > > > > > of
> > > > > > > > > all
> > > > > > > > > > I can't agree more that StateCatalog is needed and is a
> > > planned
> > > > > > > > building
> > > > > > > > > > block in an upcoming
> > > > > > > > > > FLIP but not sure how can it help now? No matter what,
> your
> > > > > > knowledge
> > > > > > > > is
> > > > > > > > > > essential when we add StateCatalog. Let me expose my
> > > > > understanding
> > > > > > in
> > > > > > > > > this
> > > > > > > > > > area:
> > > > > > > > > > * First we need create table statements to access state
> > data
> > > > and
> > > > > > > > metadata
> > > > > > > > > > * When we have that then we can add StateCatalog which
> > could
> > > > > > > > potentially
> > > > > > > > > > ease the life of users by for ex. giving off-the-shelf
> > tables
> > > > > > without
> > > > > > > > > > sweating with create table statements
> > > > > > > > > >
> > > > > > > > > > User expectations:
> > > > > > > > > > * See state data (this is fulfilled with the existing
> > > > connector)
> > > > > > > > > > * See metadata about state data like TTL (this can be
> added
> > > as
> > > > > > > metadata
> > > > > > > > > > column as you suggested since it belongs to the data)
> > > > > > > > > > * See metadata about operators (this can be added from
> > > > > > > > > savepoint-metadata)
> > > > > > > > > >
> > > > > > > > > > Important to highlight that state data table format
> differs
> > > > from
> > > > > > > state
> > > > > > > > > > metadata table format. Namely one table has rows for
> state
> > > > values
> > > > > > and
> > > > > > > > > > another has rows for operators, right?
> > > > > > > > > > I think that's the reason why you've pinpointed out that
> > the
> > > > > > > suggested
> > > > > > > > > > metadata columns are somewhat clunky.
> > > > > > > > > >
> > > > > > > > > > As a conclusion I agree to add ${state-name}_ttl metadata
> > > > column
> > > > > > > later
> > > > > > > > on
> > > > > > > > > > since it belongs to the state value and adding a new
> table
> > > type
> > > > > > (like
> > > > > > > > you
> > > > > > > > > > suggested similar to PG [1])
> > > > > > > > > > for metadata. Please see how Spark does that too [2].
> > > > > > > > > >
> > > > > > > > > > If you have better approach then please elaborate with
> more
> > > > > details
> > > > > > > and
> > > > > > > > > > help me to understand your point.
> > > > > > > > > >
> > > > > > > > > > > Up until now we've seen even in TB savepoints that the
> > > number
> > > > > of
> > > > > > > keys
> > > > > > > > > can
> > > > > > > > > > > be extremely huge but not the per key state itself.
> > > > > > > > > > > But again, this is a good feature as-is and can be
> > handled
> > > > in a
> > > > > > > > > separate
> > > > > > > > > > > jira.
> > > > > > > > > >
> > > > > > > > > > I've just created
> > > > > > https://issues.apache.org/jira/browse/FLINK-37456.
> > > > > > > > > >
> > > > > > > > > > [1]
> > > > https://www.postgresql.org/docs/current/view-pg-tables.html
> > > > > > > > > > [2]
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
> https://www.databricks.com/blog/announcing-state-reader-api-new-statestore-data-source
> > > > > > > > > >
> > > > > > > > > > BR,
> > > > > > > > > > G
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > On Tue, Mar 11, 2025 at 3:55 AM Shengkai Fang <
> > > > fskm...@gmail.com
> > > > > >
> > > > > > > > wrote:
> > > > > > > > > >
> > > > > > > > > > > Hi, Gabor. Thanks for your response.
> > > > > > > > > > >
> > > > > > > > > > > > 1. State TTL for Value Columns
> > > > > > > > > > >
> > > > > > > > > > > Thank you for addressing the limitations here.
> However, I
> > > > > believe
> > > > > > > it
> > > > > > > > > > would
> > > > > > > > > > > be beneficial to further clarify the API in this FLIP
> > > > regarding
> > > > > > how
> > > > > > > > > users
> > > > > > > > > > > can specify the TTL column.
> > > > > > > > > > >
> > > > > > > > > > > One potential approach that comes to mind is using a
> > > > > standardized
> > > > > > > > > naming
> > > > > > > > > > > convention such as ${state-name}_ttl for the metadata
> > > column
> > > > > that
> > > > > > > > > defines
> > > > > > > > > > > the TTL value. In terms of implementation, the
> > > > > > listReadableMetadata
> > > > > > > > > > > function could:
> > > > > > > > > > >
> > > > > > > > > > > 1. Read the table’s columns and configuration,
> > > > > > > > > > > 2. Extract all defined state names, and
> > > > > > > > > > > 3. Return a structured list of metadata entries
> formatted
> > > as
> > > > > > > > > > > ${state-name}_ttl.
> > > > > > > > > > >
> > > > > > > > > > > WDYT?
> > > > > > > > > > >
> > > > > > > > > > > > 2. Adding a new connector with `savepoint-metadata`
> > > > > > > > > > >
> > > > > > > > > > > Introducing a new connector type at this stage may
> > > > > unnecessarily
> > > > > > > > > > complicate
> > > > > > > > > > > the system. Given that every table already belongs to a
> > > > > Catalog,
> > > > > > > > which
> > > > > > > > > is
> > > > > > > > > > > designed to provide a Factory for building source or
> sink
> > > > > > > > connectors, I
> > > > > > > > > > > propose integrating a dedicated StateCatalog instead.
> > This
> > > > > > approach
> > > > > > > > > would
> > > > > > > > > > > allow us to:
> > > > > > > > > > >
> > > > > > > > > > > 1. Leverage the Catalog’s existing capabilities to
> manage
> > > TTL
> > > > > > > > metadata
> > > > > > > > > > > (e.g., state names and TTL logic) without duplicating
> > > > > > > functionality.
> > > > > > > > > > > 2. Provide a unified interface for connector
> > instantiation
> > > > and
> > > > > > > > metadata
> > > > > > > > > > > handling through the Catalog’s Factory pattern.
> > > > > > > > > > >
> > > > > > > > > > > Would this design decision better align with our
> > > > architecture’s
> > > > > > > > > > > extensibility and reduce redundancy?
> > > > > > > > > > >
> > > > > > > > > > > > Up until now we've seen even in TB savepoints that
> the
> > > > number
> > > > > > of
> > > > > > > > keys
> > > > > > > > > > can
> > > > > > > > > > > > be extremely huge but not the per key state itself.
> > > > > > > > > > > > But again, this is a good feature as-is and can be
> > > handled
> > > > > in a
> > > > > > > > > > separate
> > > > > > > > > > > > jira.
> > > > > > > > > > >
> > > > > > > > > > > +1 for a separate jira.
> > > > > > > > > > >
> > > > > > > > > > > Best,
> > > > > > > > > > > Shengkai
> > > > > > > > > > >
> > > > > > > > > > > Gabor Somogyi <gabor.g.somo...@gmail.com>
> 于2025年3月10日周一
> > > > > 19:05写道：
> > > > > > > > > > >
> > > > > > > > > > > > Hi Shengkai,
> > > > > > > > > > > >
> > > > > > > > > > > > Please see my comments inline.
> > > > > > > > > > > >
> > > > > > > > > > > > BR,
> > > > > > > > > > > > G
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > On Mon, Mar 3, 2025 at 7:07 AM Shengkai Fang <
> > > > > > fskm...@gmail.com>
> > > > > > > > > > wrote:
> > > > > > > > > > > >
> > > > > > > > > > > > > Hi, Gabor. Thanks for your the FLIP. I have some
> > > > questions
> > > > > > > about
> > > > > > > > > the
> > > > > > > > > > > > FLIP:
> > > > > > > > > > > > >
> > > > > > > > > > > > > 1. State TTL for Value Columns
> > > > > > > > > > > > > How can users retrieve the state TTL (Time-to-Live)
> > for
> > > > > each
> > > > > > > > value
> > > > > > > > > > > > column?
> > > > > > > > > > > > > From my understanding of the current design, it
> seems
> > > > that
> > > > > > this
> > > > > > > > > > > > > functionality is not supported. Could you clarify
> if
> > > > there
> > > > > > are
> > > > > > > > > plans
> > > > > > > > > > to
> > > > > > > > > > > > > address this limitation?
> > > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > Since the state processor API is not yet exposing
> this
> > > > > > > information
> > > > > > > > > this
> > > > > > > > > > > > would require several steps.
> > > > > > > > > > > > First, the state processor API support needs to be
> > added
> > > > > which
> > > > > > > can
> > > > > > > > be
> > > > > > > > > > > then
> > > > > > > > > > > > exposed on the SQL API.
> > > > > > > > > > > > This is definitely a future improvement which is
> useful
> > > and
> > > > > can
> > > > > > > be
> > > > > > > > > > > handled
> > > > > > > > > > > > in a separate jira.
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > > 2. Metadata Table vs. Metadata Column
> > > > > > > > > > > > > The metadata information described in the FLIP
> > appears
> > > to
> > > > > be
> > > > > > > > > intended
> > > > > > > > > > > to
> > > > > > > > > > > > > describe the state files stored at a specific
> > location.
> > > > To
> > > > > > me,
> > > > > > > > this
> > > > > > > > > > > > concept
> > > > > > > > > > > > > aligns more closely with system tables like
> pg_tables
> > > in
> > > > > > > > PostgreSQL
> > > > > > > > > > [1]
> > > > > > > > > > > > or
> > > > > > > > > > > > > the INFORMATION_SCHEMA in MySQL [2].
> > > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > Adding a new connector with `savepoint-metadata` is a
> > > > > > possibility
> > > > > > > > > where
> > > > > > > > > > > we
> > > > > > > > > > > > can create such functionality.
> > > > > > > > > > > > I'm not against that, just want to have a common
> > > agreement
> > > > > that
> > > > > > > we
> > > > > > > > > > would
> > > > > > > > > > > > like to move that direction.
> > > > > > > > > > > > (As a side note not just PG but Spark also has
> similar
> > > > > approach
> > > > > > > > and I
> > > > > > > > > > > > basically like the idea).
> > > > > > > > > > > > If we would go that direction savepoint metadata can
> be
> > > > > reached
> > > > > > > in
> > > > > > > > a
> > > > > > > > > > way
> > > > > > > > > > > > that one row would represent
> > > > > > > > > > > > an operator with it's values something like this:
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
> ┌─────────┬─────────┬─────────┬─────────┬─────────┬─────────┬─────────┬────────┐
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
> │operatorN│operatorU│operatorH│paralleli│maxParall│subtaskSt│coordinat│totalSta│
> > > > > > > > > > > > │ame │id │ash │sm │elism
> > > > > > > > > > > > │atesCount│orStateSi│tesSizeI│
> > > > > > > > > > > > │ │ │ │ │ │
> > > > > > > > > > > > │zeInBytes│nBytes │
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
> ├─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼────────┤
> > > > > > > > > > > > │Source: │datagen-s│47aee9439│2 │128 │2
> > > > > │16
> > > > > > > > > > > > │546 │
> > > > > > > > > > > > │datagen-s│ource-uid│4d6ea26e2│ │ │
> > > > │
> > > > > > > > > │
> > > > > > > > > > > > │
> > > > > > > > > > > > │ource │ │d544bef0a│ │ │
> > > > │
> > > > > > > > > │
> > > > > > > > > > > > │
> > > > > > > > > > > > │ │ │37bb5 │ │ │
> > > > │
> > > > > > > > > │
> > > > > > > > > > > > │
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
> ├─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼────────┤
> > > > > > > > > > > > │long-udf-│long-udf-│6ed3f40bf│2 │128 │2
> > > > > │0
> > > > > > > > > > │0
> > > > > > > > > > > > │
> > > > > > > > > > > > │with-mast│with-mast│f3c8dfcdf│ │ │
> > > > │
> > > > > > > > > │
> > > > > > > > > > > > │
> > > > > > > > > > > > │er-hook │er-hook-u│cb95128a1│ │ │
> > > > │
> > > > > > > > > │
> > > > > > > > > > > > │
> > > > > > > > > > > > │ │id │018f1 │ │ │
> > > > │
> > > > > > > > > │
> > > > > > > > > > > > │
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
> ├─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼────────┤
> > > > > > > > > > > > │value-pro│value-pro│ca4f5fe9a│2 │128 │2
> > > > > │0
> > > > > > > > > > > > │40726 │
> > > > > > > > > > > > │cess │cess-uid │637b656f0│ │ │
> > > > │
> > > > > > > > > │
> > > > > > > > > > > > │
> > > > > > > > > > > > │ │ │9ea78b3e7│ │ │
> > > > │
> > > > > > > > > │
> > > > > > > > > > > > │
> > > > > > > > > > > > │ │ │a15b9 │ │ │
> > > > │
> > > > > > > > > │
> > > > > > > > > > > > │
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
> ├─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼────────┤
> > > > > > > > > > > >
> > > > > > > > > > > > This table can then be joined with the actually
> > existing
> > > > > > > > `savepoint`
> > > > > > > > > > > > connector created tables based on UID hash (which is
> > > unique
> > > > > and
> > > > > > > > > always
> > > > > > > > > > > > exists).
> > > > > > > > > > > > This would mean that the already existing table would
> > > need
> > > > > > only a
> > > > > > > > > > single
> > > > > > > > > > > > metadata column which is the UID hash.
> > > > > > > > > > > > WDYT?
> > > > > > > > > > > > @zakelly, plz share your thoughts too.
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > > If we opt to use metadata columns, every record in
> > the
> > > > > table
> > > > > > > > would
> > > > > > > > > > end
> > > > > > > > > > > up
> > > > > > > > > > > > > having identical values for these columns (please
> > > correct
> > > > > me
> > > > > > if
> > > > > > > > I’m
> > > > > > > > > > > > > mistaken). On the other hand, the state connector
> > > > requires
> > > > > > > users
> > > > > > > > to
> > > > > > > > > > > > specify
> > > > > > > > > > > > > an operator UID or operator UID hash, after which
> it
> > > > > outputs
> > > > > > > > > > > user-defined
> > > > > > > > > > > > > values in its records. This approach feels somewhat
> > > > > redundant
> > > > > > > to
> > > > > > > > > me.
> > > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > If we would add a new `savepoint-metadata` connector
> > then
> > > > > this
> > > > > > > can
> > > > > > > > be
> > > > > > > > > > > > addressed.
> > > > > > > > > > > > On the other hand UID and UID hash are having
> either-or
> > > > > > > > relationship
> > > > > > > > > > from
> > > > > > > > > > > > config perspective,
> > > > > > > > > > > > so when a user provides the UID then he/she can be
> > > > interested
> > > > > > in
> > > > > > > > the
> > > > > > > > > > hash
> > > > > > > > > > > > for further calculations
> > > > > > > > > > > > (the whole Flink internals are depending on the
> hash).
> > > > > Printing
> > > > > > > out
> > > > > > > > > the
> > > > > > > > > > > > human readable UID
> > > > > > > > > > > > is an explicit requirement from the user side because
> > > > hashes
> > > > > > are
> > > > > > > > not
> > > > > > > > > > > human
> > > > > > > > > > > > readable.
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > > 3. Handling LIST and MAP States in the State
> > Connector
> > > > > > > > > > > > > I have concerns about how the current design
> handles
> > > LIST
> > > > > and
> > > > > > > MAP
> > > > > > > > > > > states.
> > > > > > > > > > > > > Specifically, the state connector uses Flink SQL’s
> > MAP
> > > > and
> > > > > > > ARRAY
> > > > > > > > > > types,
> > > > > > > > > > > > > which implies that it attempts to load entire MAP
> or
> > > LIST
> > > > > > > states
> > > > > > > > > into
> > > > > > > > > > > > > memory.
> > > > > > > > > > > > >
> > > > > > > > > > > > > However, in many real-world scenarios, these states
> > can
> > > > > grow
> > > > > > > very
> > > > > > > > > > > large.
> > > > > > > > > > > > > Typically, the state API addresses this by
> providing
> > an
> > > > > > > iterator
> > > > > > > > to
> > > > > > > > > > > > > traverse elements within the state incrementally.
> I’m
> > > > > unsure
> > > > > > > > > whether
> > > > > > > > > > > I’ve
> > > > > > > > > > > > > missed something in FLIP-496 or FLIP-512, but it
> > seems
> > > > that
> > > > > > the
> > > > > > > > > > current
> > > > > > > > > > > > > design might struggle with scalability in such
> cases.
> > > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > You see it good, the current implementation keeps
> state
> > > > for a
> > > > > > > > single
> > > > > > > > > > key
> > > > > > > > > > > in
> > > > > > > > > > > > memory.
> > > > > > > > > > > > Back in the days we've considered this potential
> issue
> > > and
> > > > > > > > concluded
> > > > > > > > > > that
> > > > > > > > > > > > this is not necessarily
> > > > > > > > > > > > needed for the initial version and can be done as a
> > later
> > > > > > > > > improvement.
> > > > > > > > > > > >
> > > > > > > > > > > > Up until now we've seen even in TB savepoints that
> the
> > > > number
> > > > > > of
> > > > > > > > keys
> > > > > > > > > > can
> > > > > > > > > > > > be extremely huge but not the per key state itself.
> > > > > > > > > > > > But again, this is a good feature as-is and can be
> > > handled
> > > > > in a
> > > > > > > > > > separate
> > > > > > > > > > > > jira.
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > > Best,
> > > > > > > > > > > > > Shengkai
> > > > > > > > > > > > >
> > > > > > > > > > > > > [1]
> > > > > > > https://www.postgresql.org/docs/current/view-pg-tables.html
> > > > > > > > > > > > > [2]
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
> https://dev.mysql.com/doc/refman/8.4/en/information-schema-tables-table.html
> > > > > > > > > > > > >
> > > > > > > > > > > > > Gabor Somogyi <gabor.g.somo...@gmail.com>
> > 于2025年3月3日周一
> > > > > > > 02:00写道：
> > > > > > > > > > > > >
> > > > > > > > > > > > > > Hi Zakelly,
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > In order to shoot for simplicity `METADATA
> VIRTUAL`
> > > as
> > > > > key
> > > > > > > > words
> > > > > > > > > > for
> > > > > > > > > > > > > > definition is the target.
> > > > > > > > > > > > > > When it's not super complex the latter can be
> added
> > > > too.
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > BR,
> > > > > > > > > > > > > > G
> > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > On Sun, Mar 2, 2025 at 3:37 PM Zakelly Lan <
> > > > > > > > > zakelly....@gmail.com>
> > > > > > > > > > > > > wrote:
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > > Hi Gabor,
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > +1 for this.
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > Will the metadata column use `METADATA VIRTUAL`
> > as
> > > > key
> > > > > > > words
> > > > > > > > > for
> > > > > > > > > > > > > > > definition, or `METADATA FROM xxx VIRTUAL` for
> > > > > renaming,
> > > > > > > just
> > > > > > > > > > like
> > > > > > > > > > > > the
> > > > > > > > > > > > > > > Kafka table?
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > Best,
> > > > > > > > > > > > > > > Zakelly
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > On Sat, Mar 1, 2025 at 1:31 PM Gabor Somogyi <
> > > > > > > > > > > > > gabor.g.somo...@gmail.com>
> > > > > > > > > > > > > > > wrote:
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > Hi All,
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > I'd like to start a discussion of FLIP-512:
> Add
> > > > meta
> > > > > > > > > > information
> > > > > > > > > > > to
> > > > > > > > > > > > > SQL
> > > > > > > > > > > > > > > > state connector [1].
> > > > > > > > > > > > > > > > Feel free to add your thoughts to make this
> > > feature
> > > > > > > better.
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > [1]
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
> https://cwiki.apache.org/confluence/display/FLINK/FLIP-512%3A+Add+meta+information+to+SQL+state+connector
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > BR,
> > > > > > > > > > > > > > > > G
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>
