Hi Jeff, Thank you for the fast answer!
100. Got it, I think I am starting to understand based on your example exposing three Kafka versions. Please let me know if the following corresponds to the correct interpretation. Taking OffsetCommitValue as an example, the idea is to bump the schema to version 4 to make the record flexible. The schema version 4 will be back-ported to versions 3.0.3, 3.1.3, 3.2.4, 3.3.3, 3.4.1 and potentially 3.5. These versions will continue to serialise the record with version 3 so that downgrades from there to earlier minor and/or major versions are possible. And, these versions will deserialize the records with version 4 so that they can support flexible records, although they do not make use of tagged fields (if any is present). Then, in a future version, 3.x, x >= 6, a tag field will be added to the record - say, topic id. Downgrade from 3.x to 3.0.3, 3.1.3, 3.2.4, 3.3.3, 3.4.1 and 3.5 will be possible because these versions will perform deserialisation with the record schema version 4, that is, supporting tagged fields. Hence, this approach is both backward and forward looking and allows to extend the scope of compatible versions for downgrades. [N] 3.x, x >= 6 --> downgradable to [I] 3.0.3, 3.1.3, 3.2.4, 3.3.3, 3.4.1, 3.5 --> downgradable to e.g. [O] 3.0.0. One note though is that a transitive downgrade from 3.x, x >= 6, to 3.0.0 via the version domain [I], will not be supported. Should it be explicitly mentioned in the KIP that downgrades from 3.0.3, 3.1.3, 3.2.4, 3.3.3, 3.4.1, 3.5 to earlier versions may not be possible (if offsets or group metadata records were generated by a higher version 3.x, x >= 6)? Or am I still misunderstanding? 101. I will try to provide an example to illustrate what I mean by version. Consider the GroupMetadataValue schema version 4 and the tagged field topicId introduced in 3.6. Let's say a new optional field needs to be added in 3.7. We will have two records version 4, one with topic id, the other with topic id and the new field. The new field is semantically optional (structurally it is always optional since it is a tagged field) but we want to make the distinction between a record generated by 3.6 and one generated by 3.7. How do we resolve the ambiguity? Thanks! Alexandre Le mar. 28 mars 2023 à 16:13, Jeff Kim <jeff....@confluent.io.invalid> a écrit : > > Hi Alexandre, > > Thank you for taking a look! > > 100. I am not sure I fully understand what you mean by forcefully adding > tagged fields. Let's say VX does not have a flexible version, > VY allows deserialization but serializes with a non-flexible version, and > VZ introduces a new tagged field. > VX upgrade to VY then downgrade back to VX works because even if group > metadata changes VY will serialize with > the highest non-flexible version. VZ to VY back to VZ also works because > even if VY serializes with a non-flexible field > VZ will be able to deserialize it as it is a supported version. Does this > answer your question? > > 101. The future versioning scheme needs to be backward compatible with > older coordinators. Wouldn't segregating into 2 versions > be incompatible? > > Thanks, > Jeff > > On Tue, Mar 28, 2023 at 5:47 AM Alexandre Dupriez < > alexandre.dupr...@gmail.com> wrote: > > > Hi Jeff, Team, > > > > Thank you for the KIP. This is a very interesting approach. I feel it > > is simpler than the described alternative although it comes with > > tradeoffs, thanks for highlighting those. If I may, I would like to > > share two naive questions. > > > > 100. The KIP mentions that records will be serialised with the highest > > non-flexible version (e.g. 3 for GroupMetadataValue and > > OffsetCommitValue) so that records can be deserialized with earlier > > versions of Kafka. I am not sure I understand correctly: is the idea > > to forcefully add tagged fields at the end of the records while > > maintaining the existing version (3 for the two record types just > > mentioned) so that they can be deserialized by existing Kafka versions > > for which the version of these record types is not known as flexible, > > while at the same time preserving the new tagged fields to new Kafka > > versions abreast of the addition of a new flexible version for these > > record types? If so, is it "bypassing" the protocol convention which > > prescribes the use of a flexible version to allow the use of tagged > > fields? > > > > 101. After the bump of the records to a new version indicated as > > flexible, the record version is expected to never change while the > > underlying tagged fields could potentially still evolve over time. One > > potential downside is that we lose the benefits of the versioning > > scheme enforced by the serde protocol. Could this become a problem in > > the future if there is ever a need to segregate two distinct > > "versions" of the appended record structure held by the tagged fields? > > > > Thanks, > > Alexandre > > > > Le jeu. 23 mars 2023 à 18:15, Jeff Kim <jeff....@confluent.io.invalid> a > > écrit : > > > > > > Hi Yi, > > > > > > > Does it mean with a flexible version, the future > > > version of these value types will stay at the version where the > > flexibility > > > is first introduced? Will there be any need to bump the version again in > > > the future? > > > > > > Yes, there will be no need to bump the version since we will only be > > adding > > > tagged fields but in the chance the version is bumped, we will > > deserialize > > > to the highest known (flexible) version which will ignore unknown tagged > > > fields. > > > > > > > To enforce the version not bumping, is it possible to have a unit test > > to > > > gate? > > > > > > Do you have some tests in mind? I don't think we can tell whether a > > version > > > was bumped or not. > > > > > > Best, > > > Jeff > > > > > > On Thu, Mar 23, 2023 at 12:07 PM Yi Ding <yd...@confluent.io.invalid> > > wrote: > > > > > > > Hi Jeff, > > > > > > > > Thanks for the update. Does it mean with a flexible version, the future > > > > version of these value types will stay at the version where the > > flexibility > > > > is first introduced? Will there be any need to bump the version again > > in > > > > the future? > > > > To enforce the version not bumping, is it possible to have a unit test > > to > > > > gate? > > > > > > > > > > > > On Wed, Mar 22, 2023 at 3:19 PM Jeff Kim <jeff....@confluent.io.invalid > > > > > > > wrote: > > > > > > > > > Hi all, > > > > > > > > > > After discussing with my colleagues, I have repurposed the KIP for a > > > > > general downgrade solution for both transaction and group > > coordinators. > > > > The > > > > > KIP no longer discusses the downgrade path but instead lays out the > > > > > foundation for future downgrade solutions. > > > > > > > > > > Link: > > > > > > > > > > > > > > > > https://cwiki.apache.org/confluence/display/KAFKA/KIP-915%3A+Txn+and+Group+Coordinator+Downgrade+Foundation > > > > > > > > > > Thanks, > > > > > Jeff > > > > > > > > > > On Mon, Mar 20, 2023 at 7:37 PM Jeff Kim <jeff....@confluent.io> > > wrote: > > > > > > > > > > > Hi David and Justine, > > > > > > > > > > > > Thank you both for the detailed feedback. > > > > > > > > > > > > David, > > > > > > > > > > > > 1. That makes sense. I revised the "Reading new fields" section > > with > > > > how > > > > > > we can downgrade to the highest known version and that this was > > > > confirmed > > > > > > via unit testing. I also attempted to dive deeper into using tagged > > > > > fields > > > > > > and the rejected alternative. Please let me know what you think. > > > > > > > > > > > > 2. Under "Restrictions and Guidelines" I updated the paragraph to > > > > clearly > > > > > > state that we want to use tagged fields across all record types > > > > > introduced > > > > > > in KIP-848 including OffsetCommitValue. > > > > > > > > > > > > 3. Would it be possible to bump the OffsetCommitValue record > > version to > > > > > > make it flexible along with the changes to parse with the highest > > known > > > > > > version? I'm not sure I understand why we cannot make both changes > > > > > together. > > > > > > > > > > > > 4. I completely missed this. Added some notes at the end of > > > > "Restrictions > > > > > > and Guidelines". Unfortunately I can't think of a solution at the > > > > moment. > > > > > > Will get back to you. > > > > > > > > > > > > 5. I have a section under "Identifying New Record Types" that > > discusses > > > > > > this: > > > > > > > We can automate the cleanup by writing tombstones when the > > > > coordinator > > > > > > reads unrecognizable records. This may add duplicate work if > > tombstones > > > > > > were already added but not yet pruned by the log cleaner. > > > > > > This is a sure way to delete any unknown record types even if the > > > > > operator > > > > > > does not follow the steps. > > > > > > > > > > > > 6. Thanks, I have expanded on the section on transactional offset > > > > > commits. > > > > > > As for log compaction, my understanding was that we can control the > > > > > process > > > > > > by forcing compaction. Is my understanding incorrect? > > > > > > > > > > > > 7. We throw an exception if ConfigResource.type is unknown which is > > > > true > > > > > > in our case because KIP-848 introduces a new GROUP resource type. > > We > > > > will > > > > > > need to add some sort of safeguard if the dynamic configs are not > > > > deleted > > > > > > before the server downgrade. I'll give you an update once I update > > the > > > > > > section. > > > > > > > > > > > > Justine, > > > > > > > > > > > > On the transactional offset commits, I have updated the KIP to > > reflect > > > > > > your points after our offline discussion (much appreciated). It > > seems > > > > > that > > > > > > the work required to abort from the server side is fairly big and > > will > > > > > > require additional investigation if we are to go down this path. > > > > > > > > > > > > As for downgrade limitations, I missed that part so thanks for the > > > > catch > > > > > > (along with David). Unfortunately, the proposed design won't allow > > > > > > downgrades even after the new record types are deleted because the > > > > > existing > > > > > > OffsetCommitValue record is not flexible and will not be able to > > parse > > > > > > the topicId tagged field. I'll think more about this and get back > > to > > > > you. > > > > > > > > > > > > Best, > > > > > > Jeff > > > > > > > > > > > > On Mon, Mar 20, 2023 at 2:16 PM Justine Olshan > > > > > > <jols...@confluent.io.invalid> wrote: > > > > > > > > > > > >> Hey Jeff and David, > > > > > >> > > > > > >> Thanks for the KIP! I was also looking into this a bit since I may > > > > want > > > > > to > > > > > >> change the record format for KIP-890 as well (finally > > implementing the > > > > > >> record change from KIP-360 to better support the epoch bump. This > > will > > > > > >> potentially be helpful for me to implement that work. > > > > > >> > > > > > >> I discussed some aspects with David offline. Namely, the > > alternative > > > > > >> solution to rewrite the records in the old format upon downgrade. > > One > > > > > plus > > > > > >> to this approach is that it should trigger compaction for all the > > > > > existing > > > > > >> (new format) records. > > > > > >> The main issue we ran into was how to handle ongoing transactions. > > > > (Ie, > > > > > >> could we abort and rewrite with the old format) I think that's > > one of > > > > > the > > > > > >> main downsides. I think it could potentially be possible to do > > this, > > > > but > > > > > >> we'd definitely need a server-side mechanism to abort and rewrite > > the > > > > > >> records. > > > > > >> > > > > > >> I think the trouble I was running into with the current solution > > is > > > > that > > > > > >> we > > > > > >> can only downgrade to a version slightly before the new group > > > > > coordinator. > > > > > >> If someone was running 3.2 and they upgrade to 3.6, but find a > > bug in > > > > > 3.4 > > > > > >> (or any version between 3.5 and the one they upgraded from), then > > they > > > > > can > > > > > >> only downgrade to 3.5. This puts us in a difficult spot. I guess > > in > > > > this > > > > > >> scenario, they will need to wait for the new format records to get > > > > > cleared > > > > > >> away before downgrading again. Is that correct? > > > > > >> > > > > > >> Thanks, > > > > > >> Justine > > > > > >> > > > > > >> On Thu, Mar 16, 2023 at 10:41 AM David Jacot > > > > > <dja...@confluent.io.invalid > > > > > >> > > > > > > >> wrote: > > > > > >> > > > > > >> > Hi Jeff, > > > > > >> > > > > > > >> > Thanks for the KIP! I am really glad that we are finally > > addressing > > > > > this > > > > > >> > gap in KIP-848. I have a few general comments. > > > > > >> > > > > > > >> > 1. Overall, I feel like the important bits are not bold enough > > in > > > > the > > > > > >> KIP. > > > > > >> > I think that it is good to explain the overall upgrade/downgrade > > > > > process > > > > > >> > and to highlight where the issues are but I think that the core > > bits > > > > > >> should > > > > > >> > give more details. For instance, we should explain why relying > > on > > > > tag > > > > > >> > fields works to ignore fields added in future releases. My > > > > > >> understanding is > > > > > >> > that it works because the buffer for the tagged fields is > > serialized > > > > > at > > > > > >> the > > > > > >> > end so reading with the old version, which is a prefix of the > > new > > > > one, > > > > > >> > works. > > > > > >> > > > > > > >> > 2. Moreover, it would be great if we could make the principle > > more > > > > > >> general. > > > > > >> > My hope is that we can keep reusing the principles introduced > > in the > > > > > >> KIP in > > > > > >> > future releases as well. For instance, let's say that we need to > > > > add a > > > > > >> new > > > > > >> > field to one of the new records introduced by KIP-848 or that we > > > > will > > > > > >> have > > > > > >> > to introduce a new record type as well. Would it work for those > > > > cases > > > > > as > > > > > >> > well? > > > > > >> > > > > > > >> > 3. Regarding enabling support for tagged fields for the > > > > > >> OffsetCommitValue > > > > > >> > record, it would be great if you could give more details on the > > > > steps > > > > > to > > > > > >> > get there in the KIP. My understanding is that we would have to > > do > > > > the > > > > > >> > following: 1) Update the code which reads the records to fail > > back > > > > to > > > > > >> the > > > > > >> > highest known version if the version stored in the log is > > unknown. > > > > > Let's > > > > > >> > say that we do this in AK 3.5. 2) We need to turn on tagged > > fields > > > > for > > > > > >> the > > > > > >> > record. I think that we can only do this in AK 3.6+. > > > > > >> > > > > > > >> > 4. I may have missed this part but we should clearly explain the > > > > > >> drawback > > > > > >> > of the proposed approach as well. Say that we enable tagged > > fields > > > > for > > > > > >> > OffsetCommitValue in AK 3.6. This means that it won't be > > possible to > > > > > >> > downgrade a cluster from 3.6 to a version earlier than 3.5. > > This is > > > > a > > > > > >> > significant limitation in my opinion because, I think, users > > don't > > > > > >> > necessarily upgrade to all versions. > > > > > >> > > > > > > >> > 5. In the proposal, it is not clear about whether the old > > software > > > > > will > > > > > >> > delete unknown records or not. It is true that new records will > > be > > > > > >> deleted > > > > > >> > when the group is downgraded but this only works if the operator > > > > > >> respects > > > > > >> > the process. > > > > > >> > > > > > > >> > 6. It would be great if we could extend the rejected > > alternative. > > > > The > > > > > >> > alternative sounds clearly better when you read it so we should > > > > really > > > > > >> > explain the reason to reject it. 1) One issue that you mention > > is > > > > that > > > > > >> the > > > > > >> > log must be compacted before downgrading and we don't really > > control > > > > > >> this > > > > > >> > process. 2) Transactions may be difficult to handle. I suppose > > that > > > > it > > > > > >> is > > > > > >> > possible to handle them though. Have you thought about this? > > > > > >> > > > > > > >> > 7. For the new dynamic configs, what happens if they are kept > > and > > > > the > > > > > >> > quorum controller is downgraded? > > > > > >> > > > > > > >> > Best, > > > > > >> > David > > > > > >> > > > > > > >> > On Thu, Mar 16, 2023 at 12:56 AM Jeff Kim > > > > > <jeff....@confluent.io.invalid > > > > > >> > > > > > > >> > wrote: > > > > > >> > > > > > > >> > > Hi folks, > > > > > >> > > > > > > > >> > > I would like to start a discussion thread for KIP-915: Next > > Gen > > > > > Group > > > > > >> > > Coordinator Downgrade Path which proposes the downgrade > > design for > > > > > the > > > > > >> > new > > > > > >> > > group coordinator introduced in KIP-848 > > > > > >> > > < > > > > > >> > > > > > > > >> > > > > > > >> > > > > > > > > > > > https://cwiki.apache.org/confluence/display/KAFKA/KIP-848%3A+The+Next+Generation+of+the+Consumer+Rebalance+Protocol > > > > > >> > > > > > > > > >> > > . > > > > > >> > > > > > > > >> > > KIP: > > > > > >> > > > > > > > >> > > > > > > > >> > > > > > > >> > > > > > > > > > > > https://cwiki.apache.org/confluence/display/KAFKA/KIP-915%3A+Next+Gen+Group+Coordinator+Downgrade+Path > > > > > >> > > > > > > > >> > > Thanks, > > > > > >> > > Jeff > > > > > >> > > > > > > > >> > > > > > > >> > > > > > > > > > > > > > > > > >
