In the last community sync, we spent a little time on this topic. For Spark support, there are currently two options under consideration:
Option 2: Separate repo for the Spark support. Use branches for supporting different Spark versions. Main branch for the latest Spark version (3.2 to begin with). Tooling needs to be built for producing regular snapshots of core Iceberg in a consumable way for this repo. Unclear if commits to core Iceberg will be tested pre-commit against Spark support; my impression is that they will not be, and the Spark support build can be broken by changes to core. A variant of option 3 (which we will simply call Option 3 going forward): Single repo, separate module (subdirectory) for each Spark version to be supported. Code duplication in each Spark module (no attempt to refactor out common code). Each module built against the specific version of Spark to be supported, producing a runtime jar built against that version. CI will test all modules. Support can be provided for only building the modules a developer cares about. More input was sought and people are encouraged to voice their preference. I lean towards Option 3. - Wing Yew ps. In the sync, as Steven Wu wrote, the question was raised if the same multi-version support strategy can be adopted across engines. Based on what Steven wrote, currently the Flink developer community's bandwidth makes supporting only a single Flink version (and focusing resources on developing new features on that version) the preferred choice. If so, then no multi-version support strategy for Flink is needed at this time. On Thu, Sep 23, 2021 at 5:26 PM Steven Wu <stevenz...@gmail.com> wrote: > During the sync meeting, people talked about if and how we can have the > same version support model across engines like Flink and Spark. I can > provide some input from the Flink side. > > Flink only supports two minor versions. E.g., right now Flink 1.13 is the > latest released version. That means only Flink 1.12 and 1.13 are supported. > Feature changes or bug fixes will only be backported to 1.12 and 1.13, > unless it is a serious bug (like security). With that context, personally I > like option 1 (with one actively supported Flink version in master branch) > for the iceberg-flink module. > > We discussed the idea of supporting multiple Flink versions via shm layer > and multiple modules. While it may be a little better to support multiple > Flink versions, I don't know if there is enough support and resources from > the community to pull it off. Also the ongoing maintenance burden for each > minor version release from Flink, which happens roughly every 4 months. > > > On Thu, Sep 16, 2021 at 10:25 PM Peter Vary <pv...@cloudera.com.invalid> > wrote: > >> Since you mentioned Hive, I chime in with what we do there. You might >> find it useful: >> - metastore module - only small differences - DynConstructor solves for us >> - mr module - some bigger differences, but still manageable for Hive 2-3. >> Need some new classes, but most of the code is reused - extra module for >> Hive 3. For Hive 4 we use a different repo as we moved to the Hive >> codebase. >> >> My thoughts based on the above experience: >> - Keeping Hive 4 and Hive 2-3 code in sync is a pain. We constantly have >> problems with backporting changes between repos and we are slacking behind >> which hurts both projects >> - Hive 2-3 model is working better by forcing us to keep the things in >> sync, but with serious differences in the Hive project it still doesn't >> seem like a viable option. >> >> So I think the question is: How stable is the Spark code we are >> integrating to. If I is fairly stable then we are better off with a "one >> repo multiple modules" approach and we should consider the multirepo only >> if the differences become prohibitive. >> >> Thanks, Peter >> >> On Fri, 17 Sep 2021, 02:21 Anton Okolnychyi, >> <aokolnyc...@apple.com.invalid> wrote: >> >>> Okay, looks like there is consensus around supporting multiple Spark >>> versions at the same time. There are folks who mentioned this on this >>> thread and there were folks who brought this up during the sync. >>> >>> Let’s think through Option 2 and 3 in more detail then. >>> >>> Option 2 >>> >>> In Option 2, there will be a separate repo. I believe the master branch >>> will soon point to Spark 3.2 (the most recent supported version). The main >>> development will happen there and the artifact version will be 0.1.0. I >>> also suppose there will be 0.1.x-spark-2 and 0.1.x-spark-3.1 branches where >>> we will cherry-pick applicable changes. Once we are ready to release 0.1.0 >>> Spark integration, we will create 0.1.x-spark-3.2 and cut 3 releases: Spark >>> 2.4, Spark 3.1, Spark 3.2. After that, we will bump the version in master >>> to 0.2.0 and create new 0.2.x-spark-2 and 0.2.x-spark-3.1 branches for >>> cherry-picks. >>> >>> I guess we will continue to shade everything in the new repo and will >>> have to release every time the core is released. We will do a maintenance >>> release for each supported Spark version whenever we cut a new maintenance >>> Iceberg >>> release or need to fix any bugs in the Spark integration. >>> Under this model, we will probably need nightly snapshots (or on each >>> commit) for the core format and the Spark integration will depend on >>> snapshots until we are ready to release. >>> >>> Overall, I think this option gives us very simple builds and provides >>> best separation. It will keep the main repo clean. The main downside is >>> that we will have to split a Spark feature into two PRs: one against the >>> core and one against the Spark integration. Certain changes in core can >>> also break the Spark integration too and will require adaptations. >>> >>> Ryan, I am not sure I fully understood the testing part. How will we be >>> able to test the Spark integration in the main repo if certain changes in >>> core may break the Spark integration and require changes there? Will we try >>> to prohibit such changes? >>> >>> Option 3 (modified) >>> >>> If I get correctly, the modified Option 3 sounds very close to >>> the initially suggested approach by Imran but with code duplication instead >>> of extra refactoring and introducing new common modules. >>> >>> Jack, are you suggesting we test only a single Spark version at a time? >>> Or do we expect to test all versions? Will there be any difference compared >>> to just having a module per version? I did not fully understand. >>> >>> My worry with this approach is that our build will be very complicated >>> and we will still have a lot of Spark-related modules in the main repo. >>> Once people start using Flink and Hive more, will we have to do the same? >>> >>> - Anton >>> >>> >>> >>> On 16 Sep 2021, at 08:11, Ryan Blue <b...@tabular.io> wrote: >>> >>> I'd support the option that Jack suggests if we can set a few >>> expectations for keeping it clean. >>> >>> First, I'd like to avoid refactoring code to share it across Spark >>> versions -- that introduces risk because we're relying on compiling against >>> one version and running in another and both Spark and Scala change rapidly. >>> A big benefit of options 1 and 2 is that we mostly focus on only one Spark >>> version. I think we should duplicate code rather than spend time >>> refactoring to rely on binary compatibility. I propose we start each new >>> Spark version by copying the last one and updating it. And we should build >>> just the latest supported version by default. >>> >>> The drawback to having everything in a single repo is that we wouldn't >>> be able to cherry-pick changes across Spark versions/branches, but I think >>> Jack is right that having a single build is better. >>> >>> Second, we should make CI faster by running the Spark builds in >>> parallel. It sounds like this is what would happen anyway, with a property >>> that selects the Spark version that you want to build against. >>> >>> Overall, this new suggestion sounds like a promising way forward. >>> >>> Ryan >>> >>> On Wed, Sep 15, 2021 at 11:46 PM Jack Ye <yezhao...@gmail.com> wrote: >>> >>>> I think in Ryan's proposal we will create a ton of modules anyway, as >>>> Wing listed we are just using git branch as an additional dimension, but my >>>> understanding is that you will still have 1 core, 1 extension, 1 runtime >>>> artifact published for each Spark version in either approach. >>>> >>>> In that case, this is just brainstorming, I wonder if we can explore a >>>> modified option 3 that flattens all the versions in each Spark branch in >>>> option 2 into master. The repository structure would look something like: >>>> >>>> iceberg/api/... >>>> /bundled-guava/... >>>> /core/... >>>> ... >>>> /spark/2.4/core/... >>>> /extension/... >>>> /runtime/... >>>> /3.1/core/... >>>> /extension/... >>>> /runtime/... >>>> >>>> The gradle build script in the root is configured to build against the >>>> latest version of Spark by default, unless otherwise specified by the user. >>>> >>>> Intellij can also be configured to only index files of specific >>>> versions based on the same config used in build. >>>> >>>> In this way, I imagine the CI setup to be much easier to do things like >>>> testing version compatibility for a feature or running only a >>>> specific subset of Spark version builds based on the Spark version >>>> directories touched. >>>> >>>> And the biggest benefit is that we don't have the same difficulty as >>>> option 2 of developing a feature when it's both in core and Spark. >>>> >>>> We can then develop a mechanism to vote to stop support of certain >>>> versions, and archive the corresponding directory to avoid accumulating too >>>> many versions in the long term. >>>> >>>> -Jack Ye >>>> >>>> >>>> On Wed, Sep 15, 2021 at 4:17 PM Ryan Blue <b...@tabular.io> wrote: >>>> >>>>> Sorry, I was thinking about CI integration between Iceberg Java and >>>>> Iceberg Spark, I just didn't mention it and I see how that's a big thing >>>>> to >>>>> leave out! >>>>> >>>>> I would definitely want to test the projects together. One thing we >>>>> could do is have a nightly build like Russell suggests. I'm also wondering >>>>> if we could have some tighter integration where the Iceberg Spark build >>>>> can >>>>> be included in the Iceberg Java build using properties. Maybe the github >>>>> action could checkout Iceberg, then checkout the Spark integration's >>>>> latest >>>>> branch, and then run the gradle build with a property that makes Spark a >>>>> subproject in the build. That way we can continue to have Spark CI run >>>>> regularly. >>>>> >>>>> On Wed, Sep 15, 2021 at 3:08 PM Russell Spitzer < >>>>> russell.spit...@gmail.com> wrote: >>>>> >>>>>> I agree that Option 2 is considerably more difficult for development >>>>>> when core API changes need to be picked up by the external Spark module. >>>>>> I >>>>>> also think a monthly release would probably still be prohibitive to >>>>>> actually implementing new features that appear in the API, I would hope >>>>>> we >>>>>> have a much faster process or maybe just have snapshot artifacts >>>>>> published >>>>>> nightly? >>>>>> >>>>>> On Sep 15, 2021, at 4:46 PM, Wing Yew Poon < >>>>>> wyp...@cloudera.com.INVALID> wrote: >>>>>> >>>>>> IIUC, Option 2 is to move the Spark support for Iceberg into a >>>>>> separate repo (subproject of Iceberg). Would we have branches such as >>>>>> 0.13-2.4, 0.13-3.0, 0.13-3.1, and 0.13-3.2? For features that can be >>>>>> supported in all versions or all Spark 3 versions, then we would need to >>>>>> commit the changes to all applicable branches. Basically we are trading >>>>>> more work to commit to multiple branches for simplified build and CI >>>>>> time per branch, which might be an acceptable trade-off. However, the >>>>>> biggest downside is that changes may need to be made in core Iceberg as >>>>>> well as in the engine (in this case Spark) support, and we need to wait >>>>>> for >>>>>> a release of core Iceberg to consume the changes in the subproject. In >>>>>> this >>>>>> case, maybe we should have a monthly release of core Iceberg (no matter >>>>>> how >>>>>> many changes go in, as long as it is non-zero) so that the subproject can >>>>>> consume changes fairly quickly? >>>>>> >>>>>> >>>>>> On Wed, Sep 15, 2021 at 2:09 PM Ryan Blue <b...@tabular.io> wrote: >>>>>> >>>>>>> Thanks for bringing this up, Anton. I’m glad that we have the set of >>>>>>> potential solutions well defined. >>>>>>> >>>>>>> Looks like the next step is to decide whether we want to require >>>>>>> people to update Spark versions to pick up newer versions of Iceberg. >>>>>>> If we >>>>>>> choose to make people upgrade, then option 1 is clearly the best choice. >>>>>>> >>>>>>> I don’t think that we should make updating Spark a requirement. Many >>>>>>> of the things that we’re working on are orthogonal to Spark versions, >>>>>>> like >>>>>>> table maintenance actions, secondary indexes, the 1.0 API, views, ORC >>>>>>> delete files, new storage implementations, etc. Upgrading Spark is time >>>>>>> consuming and untrusted in my experience, so I think we would be >>>>>>> setting up >>>>>>> an unnecessary trade-off between spending lots of time to upgrade Spark >>>>>>> and >>>>>>> picking up new Iceberg features. >>>>>>> >>>>>>> Another way of thinking about this is that if we went with option 1, >>>>>>> then we could port bug fixes into 0.12.x. But there are many things that >>>>>>> wouldn’t fit this model, like adding a FileIO implementation for ADLS. >>>>>>> So >>>>>>> some people in the community would have to maintain branches of newer >>>>>>> Iceberg versions with older versions of Spark outside of the main >>>>>>> Iceberg >>>>>>> project — that defeats the purpose of simplifying things with option 1 >>>>>>> because we would then have more people maintaining the same 0.13.x with >>>>>>> Spark 3.1 branch. (This reminds me of the Spark community, where we >>>>>>> wanted >>>>>>> to release a 2.5 line with DSv2 backported, but the community decided >>>>>>> not >>>>>>> to so we built similar 2.4+DSv2 branches at Netflix, Tencent, Apple, >>>>>>> etc.) >>>>>>> >>>>>>> If the community is going to do the work anyway — and I think some >>>>>>> of us would — we should make it possible to share that work. That’s why >>>>>>> I >>>>>>> don’t think that we should go with option 1. >>>>>>> >>>>>>> If we don’t go with option 1, then the choice is how to maintain >>>>>>> multiple Spark versions. I think that the way we’re doing it right now >>>>>>> is >>>>>>> not something we want to continue. >>>>>>> >>>>>>> Using multiple modules (option 3) is concerning to me because of the >>>>>>> changes in Spark. We currently structure the library to share as much >>>>>>> code >>>>>>> as possible. But that means compiling against different Spark versions >>>>>>> and >>>>>>> relying on binary compatibility and reflection in some cases. To me, >>>>>>> this >>>>>>> seems unmaintainable in the long run because it requires refactoring >>>>>>> common >>>>>>> classes and spending a lot of time deduplicating code. It also creates a >>>>>>> ton of modules, at least one common module, then a module per version, >>>>>>> then >>>>>>> an extensions module per version, and finally a runtime module per >>>>>>> version. >>>>>>> That’s 3 modules per Spark version, plus any new common modules. And >>>>>>> each >>>>>>> module needs to be tested, which is making our CI take a really long >>>>>>> time. >>>>>>> We also don’t support multiple Scala versions, which is another gap that >>>>>>> will require even more modules and tests. >>>>>>> >>>>>>> I like option 2 because it would allow us to compile against a >>>>>>> single version of Spark (which will be much more reliable). It would >>>>>>> give >>>>>>> us an opportunity to support different Scala versions. It avoids the >>>>>>> need >>>>>>> to refactor to share code and allows people to focus on a single >>>>>>> version of >>>>>>> Spark, while also creating a way for people to maintain and update the >>>>>>> older versions with newer Iceberg releases. I don’t think that this >>>>>>> would >>>>>>> slow down development. I think it would actually speed it up because >>>>>>> we’d >>>>>>> be spending less time trying to make multiple versions work in the same >>>>>>> build. And anyone in favor of option 1 would basically get option 1: you >>>>>>> don’t have to care about branches for older Spark versions. >>>>>>> >>>>>>> Jack makes a good point about wanting to keep code in a single >>>>>>> repository, but I think that the need to manage more version >>>>>>> combinations >>>>>>> overrides this concern. It’s easier to make this decision in python >>>>>>> because >>>>>>> we’re not trying to depend on two projects that change relatively >>>>>>> quickly. >>>>>>> We’re just trying to build a library. >>>>>>> >>>>>>> Ryan >>>>>>> >>>>>>> On Wed, Sep 15, 2021 at 2:58 AM OpenInx <open...@gmail.com> wrote: >>>>>>> >>>>>>>> Thanks for bringing this up, Anton. >>>>>>>> >>>>>>>> Everyone has great pros/cons to support their preferences. Before >>>>>>>> giving my preference, let me raise one question: what's the top >>>>>>>> priority >>>>>>>> thing for apache iceberg project at this point in time ? This question >>>>>>>> will help us to answer the following question: Should we support more >>>>>>>> engine versions more robustly or be a bit more aggressive and >>>>>>>> concentrate >>>>>>>> on getting the new features that users need most in order to keep the >>>>>>>> project more competitive ? >>>>>>>> >>>>>>>> If people watch the apache iceberg project and check the issues & >>>>>>>> PR frequently, I guess more than 90% people will answer the priority >>>>>>>> question: There is no doubt for making the whole v2 story to be >>>>>>>> production-ready. The current roadmap discussion also proofs the >>>>>>>> thing : >>>>>>>> https://lists.apache.org/x/thread.html/r84e80216c259c81f824c6971504c321cd8c785774c489d52d4fc123f@%3Cdev.iceberg.apache.org%3E >>>>>>>> . >>>>>>>> >>>>>>>> In order to ensure the highest priority at this point in time, I >>>>>>>> will prefer option-1 to reduce the cost of engine maintenance, so as to >>>>>>>> free up resources to make v2 production-ready. >>>>>>>> >>>>>>>> On Wed, Sep 15, 2021 at 3:00 PM Saisai Shao <sai.sai.s...@gmail.com> >>>>>>>> wrote: >>>>>>>> >>>>>>>>> From Dev's point, it has less burden to always support the latest >>>>>>>>> version of Spark (for example). But from user's point, especially for >>>>>>>>> us >>>>>>>>> who maintain Spark internally, it is not easy to upgrade the Spark >>>>>>>>> version >>>>>>>>> for the first time (since we have many customizations internally), and >>>>>>>>> we're still promoting to upgrade to 3.1.2. If the community ditches >>>>>>>>> the >>>>>>>>> support of old version of Spark3, users have to maintain it themselves >>>>>>>>> unavoidably. >>>>>>>>> >>>>>>>>> So I'm inclined to make this support in community, not by users >>>>>>>>> themselves, as for Option 2 or 3, I'm fine with either. And to >>>>>>>>> relieve the >>>>>>>>> burden, we could support limited versions of Spark (for example 2 >>>>>>>>> versions). >>>>>>>>> >>>>>>>>> Just my two cents. >>>>>>>>> >>>>>>>>> -Saisai >>>>>>>>> >>>>>>>>> >>>>>>>>> Jack Ye <yezhao...@gmail.com> 于2021年9月15日周三 下午1:35写道： >>>>>>>>> >>>>>>>>>> Hi Wing Yew, >>>>>>>>>> >>>>>>>>>> I think 2.4 is a different story, we will continue to support >>>>>>>>>> Spark 2.4, but as you can see it will continue to have very limited >>>>>>>>>> functionalities comparing to Spark 3. I believe we discussed about >>>>>>>>>> option 3 >>>>>>>>>> when we were doing Spark 3.0 to 3.1 upgrade. Recently we are seeing >>>>>>>>>> the >>>>>>>>>> same issue for Flink 1.11, 1.12 and 1.13 as well. I feel we need a >>>>>>>>>> consistent strategy around this, let's take this chance to make a >>>>>>>>>> good >>>>>>>>>> community guideline for all future engine versions, especially for >>>>>>>>>> Spark, >>>>>>>>>> Flink and Hive that are in the same repository. >>>>>>>>>> >>>>>>>>>> I can totally understand your point of view Wing, in fact, >>>>>>>>>> speaking from the perspective of AWS EMR, we have to support over 40 >>>>>>>>>> versions of the software because there are people who are still >>>>>>>>>> using Spark >>>>>>>>>> 1.4, believe it or not. After all, keep backporting changes will >>>>>>>>>> become a >>>>>>>>>> liability not only on the user side, but also on the service >>>>>>>>>> provider side, >>>>>>>>>> so I believe it's not a bad practice to push for user upgrade, as it >>>>>>>>>> will >>>>>>>>>> make the life of both parties easier in the end. New feature is >>>>>>>>>> definitely >>>>>>>>>> one of the best incentives to promote an upgrade on user side. >>>>>>>>>> >>>>>>>>>> I think the biggest issue of option 3 is about its scalability, >>>>>>>>>> because we will have an unbounded list of packages to add and >>>>>>>>>> compile in >>>>>>>>>> the future, and we probably cannot drop support of that package once >>>>>>>>>> created. If we go with option 1, I think we can still publish a few >>>>>>>>>> patch >>>>>>>>>> versions for old Iceberg releases, and committers can control the >>>>>>>>>> amount of >>>>>>>>>> patch versions to guard people from abusing the power of patching. I >>>>>>>>>> see >>>>>>>>>> this as a consistent strategy also for Flink and Hive. With this >>>>>>>>>> strategy, >>>>>>>>>> we can truly have a compatibility matrix for engine versions against >>>>>>>>>> Iceberg versions. >>>>>>>>>> >>>>>>>>>> -Jack >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> On Tue, Sep 14, 2021 at 10:00 PM Wing Yew Poon < >>>>>>>>>> wyp...@cloudera.com.invalid> wrote: >>>>>>>>>> >>>>>>>>>>> I understand and sympathize with the desire to use new DSv2 >>>>>>>>>>> features in Spark 3.2. I agree that Option 1 is the easiest for >>>>>>>>>>> developers, >>>>>>>>>>> but I don't think it considers the interests of users. I do not >>>>>>>>>>> think that >>>>>>>>>>> most users will upgrade to Spark 3.2 as soon as it is released. It >>>>>>>>>>> is a >>>>>>>>>>> "minor version" upgrade in name from 3.1 (or from 3.0), but I think >>>>>>>>>>> we all >>>>>>>>>>> know that it is not a minor upgrade. There are a lot of changes >>>>>>>>>>> from 3.0 to >>>>>>>>>>> 3.1 and from 3.1 to 3.2. I think there are even a lot of users >>>>>>>>>>> running >>>>>>>>>>> Spark 2.4 and not even on Spark 3 yet. Do we also plan to stop >>>>>>>>>>> supporting >>>>>>>>>>> Spark 2.4? >>>>>>>>>>> >>>>>>>>>>> Please correct me if I'm mistaken, but the folks who have spoken >>>>>>>>>>> out in favor of Option 1 all work for the same organization, don't >>>>>>>>>>> they? >>>>>>>>>>> And they don't have a problem with making their users, all >>>>>>>>>>> internal, simply >>>>>>>>>>> upgrade to Spark 3.2, do they? (Or they are already running an >>>>>>>>>>> internal >>>>>>>>>>> fork that is close to 3.2.) >>>>>>>>>>> >>>>>>>>>>> I work for an organization with customers running different >>>>>>>>>>> versions of Spark. It is true that we can backport new features to >>>>>>>>>>> older >>>>>>>>>>> versions if we wanted to. I suppose the people contributing to >>>>>>>>>>> Iceberg work >>>>>>>>>>> for some organization or other that either use Iceberg in-house, or >>>>>>>>>>> provide >>>>>>>>>>> software (possibly in the form of a service) to customers, and >>>>>>>>>>> either way, >>>>>>>>>>> the organizations have the ability to backport features and fixes to >>>>>>>>>>> internal versions. Are there any users out there who simply use >>>>>>>>>>> Apache >>>>>>>>>>> Iceberg and depend on the community version? >>>>>>>>>>> >>>>>>>>>>> There may be features that are broadly useful that do not depend >>>>>>>>>>> on Spark 3.2. Is it worth supporting them on Spark 3.0/3.1 (and >>>>>>>>>>> even 2.4)? >>>>>>>>>>> >>>>>>>>>>> I am not in favor of Option 2. I do not oppose Option 1, but I >>>>>>>>>>> would consider Option 3 too. Anton, you said 5 modules are >>>>>>>>>>> required; what >>>>>>>>>>> are the modules you're thinking of? >>>>>>>>>>> >>>>>>>>>>> - Wing Yew >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> On Tue, Sep 14, 2021 at 5:38 PM Yufei Gu <flyrain...@gmail.com> >>>>>>>>>>> wrote: >>>>>>>>>>> >>>>>>>>>>>> Option 1 sounds good to me. Here are my reasons: >>>>>>>>>>>> >>>>>>>>>>>> 1. Both 2 and 3 will slow down the development. Considering the >>>>>>>>>>>> limited resources in the open source community, the upsides of >>>>>>>>>>>> option 2 and >>>>>>>>>>>> 3 are probably not worthy. >>>>>>>>>>>> 2. Both 2 and 3 assume the use cases may not exist. It's hard >>>>>>>>>>>> to predict anything, but even if these use cases are legit, users >>>>>>>>>>>> can still >>>>>>>>>>>> get the new feature by backporting it to an older version in case >>>>>>>>>>>> of >>>>>>>>>>>> upgrading to a newer version isn't an option. >>>>>>>>>>>> >>>>>>>>>>>> Best, >>>>>>>>>>>> >>>>>>>>>>>> Yufei >>>>>>>>>>>> >>>>>>>>>>>> `This is not a contribution` >>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>>> On Tue, Sep 14, 2021 at 4:54 PM Anton Okolnychyi < >>>>>>>>>>>> aokolnyc...@apple.com.invalid> wrote: >>>>>>>>>>>> >>>>>>>>>>>>> To sum up what we have so far: >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> *Option 1 (support just the most recent minor Spark 3 version)* >>>>>>>>>>>>> >>>>>>>>>>>>> The easiest option for us devs, forces the user to upgrade to >>>>>>>>>>>>> the most recent minor Spark version to consume any new >>>>>>>>>>>>> Iceberg features. >>>>>>>>>>>>> >>>>>>>>>>>>> *Option 2 (a separate project under Iceberg)* >>>>>>>>>>>>> >>>>>>>>>>>>> Can support as many Spark versions as needed and the codebase >>>>>>>>>>>>> is still separate as we can use separate branches. >>>>>>>>>>>>> Impossible to consume any unreleased changes in core, may slow >>>>>>>>>>>>> down the development. >>>>>>>>>>>>> >>>>>>>>>>>>> *Option 3 (separate modules for Spark 3.1/3.2)* >>>>>>>>>>>>> >>>>>>>>>>>>> Introduce more modules in the same project. >>>>>>>>>>>>> Can consume unreleased changes but it will required at least 5 >>>>>>>>>>>>> modules to support 2.4, 3.1 and 3.2, making the build and testing >>>>>>>>>>>>> complicated. >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> Are there any users for whom upgrading the minor Spark version >>>>>>>>>>>>> (e3.1 to 3.2) to consume new features is a blocker? >>>>>>>>>>>>> We follow Option 1 internally at the moment but I would like >>>>>>>>>>>>> to hear what other people think/need. >>>>>>>>>>>>> >>>>>>>>>>>>> - Anton >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> On 14 Sep 2021, at 09:44, Russell Spitzer < >>>>>>>>>>>>> russell.spit...@gmail.com> wrote: >>>>>>>>>>>>> >>>>>>>>>>>>> I think we should go for option 1. I already am not a big fan >>>>>>>>>>>>> of having runtime errors for unsupported things based on versions >>>>>>>>>>>>> and I >>>>>>>>>>>>> don't think minor version upgrades are a large issue for users. >>>>>>>>>>>>> I'm >>>>>>>>>>>>> especially not looking forward to supporting interfaces that only >>>>>>>>>>>>> exist in >>>>>>>>>>>>> Spark 3.2 in a multiple Spark version support future. >>>>>>>>>>>>> >>>>>>>>>>>>> On Sep 14, 2021, at 11:32 AM, Anton Okolnychyi < >>>>>>>>>>>>> aokolnyc...@apple.com.INVALID> wrote: >>>>>>>>>>>>> >>>>>>>>>>>>> First of all, is option 2 a viable option? We discussed >>>>>>>>>>>>> separating the python module outside of the project a few weeks >>>>>>>>>>>>> ago, and >>>>>>>>>>>>> decided to not do that because it's beneficial for code cross >>>>>>>>>>>>> reference and >>>>>>>>>>>>> more intuitive for new developers to see everything in the same >>>>>>>>>>>>> repository. >>>>>>>>>>>>> I would expect the same argument to also hold here. >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> That’s exactly the concern I have about Option 2 at this >>>>>>>>>>>>> moment. >>>>>>>>>>>>> >>>>>>>>>>>>> Overall I would personally prefer us to not support all the >>>>>>>>>>>>> minor versions, but instead support maybe just 2-3 latest >>>>>>>>>>>>> versions in a >>>>>>>>>>>>> major version. >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> This is when it gets a bit complicated. If we want to support >>>>>>>>>>>>> both Spark 3.1 and Spark 3.2 with a single module, it means we >>>>>>>>>>>>> have to >>>>>>>>>>>>> compile against 3.1. The problem is that we rely on DSv2 that is >>>>>>>>>>>>> being >>>>>>>>>>>>> actively developed. 3.2 and 3.1 have substantial differences. On >>>>>>>>>>>>> top of >>>>>>>>>>>>> that, we have our extensions that are extremely low-level and may >>>>>>>>>>>>> break not >>>>>>>>>>>>> only between minor versions but also between patch releases. >>>>>>>>>>>>> >>>>>>>>>>>>> f there are some features requiring a newer version, it makes >>>>>>>>>>>>> sense to move that newer version in master. >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> Internally, we don’t deliver new features to older Spark >>>>>>>>>>>>> versions as it requires a lot of effort to port things. >>>>>>>>>>>>> Personally, I don’t >>>>>>>>>>>>> think it is too bad to require users to upgrade if they want new >>>>>>>>>>>>> features. >>>>>>>>>>>>> At the same time, there are valid concerns with this approach too >>>>>>>>>>>>> that we >>>>>>>>>>>>> mentioned during the sync. For example, certain new features >>>>>>>>>>>>> would also >>>>>>>>>>>>> work fine with older Spark versions. I generally agree with that >>>>>>>>>>>>> and that >>>>>>>>>>>>> not supporting recent versions is not ideal. However, I want to >>>>>>>>>>>>> find a >>>>>>>>>>>>> balance between the complexity on our side and ease of use for >>>>>>>>>>>>> the users. >>>>>>>>>>>>> Ideally, supporting a few recent versions would be sufficient but >>>>>>>>>>>>> our Spark >>>>>>>>>>>>> integration is too low-level to do that with a single module. >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> On 13 Sep 2021, at 20:53, Jack Ye <yezhao...@gmail.com> wrote: >>>>>>>>>>>>> >>>>>>>>>>>>> First of all, is option 2 a viable option? We discussed >>>>>>>>>>>>> separating the python module outside of the project a few weeks >>>>>>>>>>>>> ago, and >>>>>>>>>>>>> decided to not do that because it's beneficial for code cross >>>>>>>>>>>>> reference and >>>>>>>>>>>>> more intuitive for new developers to see everything in the same >>>>>>>>>>>>> repository. >>>>>>>>>>>>> I would expect the same argument to also hold here. >>>>>>>>>>>>> >>>>>>>>>>>>> Overall I would personally prefer us to not support all the >>>>>>>>>>>>> minor versions, but instead support maybe just 2-3 latest >>>>>>>>>>>>> versions in a >>>>>>>>>>>>> major version. This avoids the problem that some users are >>>>>>>>>>>>> unwilling to >>>>>>>>>>>>> move to a newer version and keep patching old Spark version >>>>>>>>>>>>> branches. If >>>>>>>>>>>>> there are some features requiring a newer version, it makes sense >>>>>>>>>>>>> to move >>>>>>>>>>>>> that newer version in master. >>>>>>>>>>>>> >>>>>>>>>>>>> In addition, because currently Spark is considered the most >>>>>>>>>>>>> feature-complete reference implementation compared to all other >>>>>>>>>>>>> engines, I >>>>>>>>>>>>> think we should not add artificial barriers that would slow down >>>>>>>>>>>>> its >>>>>>>>>>>>> development speed. >>>>>>>>>>>>> >>>>>>>>>>>>> So my thinking is closer to option 1. >>>>>>>>>>>>> >>>>>>>>>>>>> Best, >>>>>>>>>>>>> Jack Ye >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> On Mon, Sep 13, 2021 at 7:39 PM Anton Okolnychyi < >>>>>>>>>>>>> aokolnyc...@apple.com.invalid> wrote: >>>>>>>>>>>>> >>>>>>>>>>>>>> Hey folks, >>>>>>>>>>>>>> >>>>>>>>>>>>>> I want to discuss our Spark version support strategy. >>>>>>>>>>>>>> >>>>>>>>>>>>>> So far, we have tried to support both 3.0 and 3.1. It is >>>>>>>>>>>>>> great to support older versions but because we compile against >>>>>>>>>>>>>> 3.0, we >>>>>>>>>>>>>> cannot use any Spark features that are offered in newer versions. >>>>>>>>>>>>>> Spark 3.2 is just around the corner and it brings a lot of >>>>>>>>>>>>>> important features such dynamic filtering for v2 tables, required >>>>>>>>>>>>>> distribution and ordering for writes, etc. These features are >>>>>>>>>>>>>> too important >>>>>>>>>>>>>> to ignore them. >>>>>>>>>>>>>> >>>>>>>>>>>>>> Apart from that, I have an end-to-end prototype for >>>>>>>>>>>>>> merge-on-read with Spark that actually leverages some of the 3.2 >>>>>>>>>>>>>> features. >>>>>>>>>>>>>> I’ll be implementing all new Spark DSv2 APIs for us internally >>>>>>>>>>>>>> and would >>>>>>>>>>>>>> love to share that with the rest of the community. >>>>>>>>>>>>>> >>>>>>>>>>>>>> I see two options to move forward: >>>>>>>>>>>>>> >>>>>>>>>>>>>> Option 1 >>>>>>>>>>>>>> >>>>>>>>>>>>>> Migrate to Spark 3.2 in master, maintain 0.12 for a while by >>>>>>>>>>>>>> releasing minor versions with bug fixes. >>>>>>>>>>>>>> >>>>>>>>>>>>>> Pros: almost no changes to the build configuration, no extra >>>>>>>>>>>>>> work on our side as just a single Spark version is actively >>>>>>>>>>>>>> maintained. >>>>>>>>>>>>>> Cons: some new features that we will be adding to master >>>>>>>>>>>>>> could also work with older Spark versions but all 0.12 releases >>>>>>>>>>>>>> will only >>>>>>>>>>>>>> contain bug fixes. Therefore, users will be forced to migrate to >>>>>>>>>>>>>> Spark 3.2 >>>>>>>>>>>>>> to consume any new Spark or format features. >>>>>>>>>>>>>> >>>>>>>>>>>>>> Option 2 >>>>>>>>>>>>>> >>>>>>>>>>>>>> Move our Spark integration into a separate project and >>>>>>>>>>>>>> introduce branches for 3.0, 3.1 and 3.2. >>>>>>>>>>>>>> >>>>>>>>>>>>>> Pros: decouples the format version from Spark, we can support >>>>>>>>>>>>>> as many Spark versions as needed. >>>>>>>>>>>>>> Cons: more work initially to set everything up, more work to >>>>>>>>>>>>>> release, will need a new release of the core format to consume >>>>>>>>>>>>>> any changes >>>>>>>>>>>>>> in the Spark integration. >>>>>>>>>>>>>> >>>>>>>>>>>>>> Overall, I think option 2 seems better for the user but my >>>>>>>>>>>>>> main worry is that we will have to release the format more >>>>>>>>>>>>>> frequently >>>>>>>>>>>>>> (which is a good thing but requires more work and time) and the >>>>>>>>>>>>>> overall >>>>>>>>>>>>>> Spark development may be slower. >>>>>>>>>>>>>> >>>>>>>>>>>>>> I’d love to hear what everybody thinks about this matter. >>>>>>>>>>>>>> >>>>>>>>>>>>>> Thanks, >>>>>>>>>>>>>> Anton >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>> >>>>>>> -- >>>>>>> Ryan Blue >>>>>>> Tabular >>>>>>> >>>>>> >>>>>> >>>>> >>>>> -- >>>>> Ryan Blue >>>>> Tabular >>>>> >>>> >>> >>> -- >>> Ryan Blue >>> Tabular >>> >>> >>>
