I agree that having both modes and let the user choose the one he/she wants is the best option (I don't see big arguments on this honestly). Once we have this, I don't see big differences on what is the default. What - I think - we still have to work on, is to go ahead with the "strict mode" work and provide a more convenient way for users to switch among the 2 options. I mean: currently we have one flag for throwing exception on overflow for operations on decimals, one for doing the same for operations on other data types and probably going ahead we will have more. I think in the end we will need to collect them all under an "umbrella" flag which lets the user simply switch between strict and non-strict mode. I also think that we will need to document this very well and give it particular attention in our docs, maybe with a dedicated section, in order to provide enough visibility on it to end users.
I’m +1 on adding a strict mode flag this way, but I’m undecided on whether or not we want a separate flag for each of the arithmetic overflow situations that could produce invalid results. My intuition is yes, because different users have different levels of tolerance for different kinds of errors. I’d expect these sorts of configurations to be set up at an infrastructure level, e.g. to maintain consistent standards throughout a whole organization. From: Gengliang Wang <gengliang.w...@databricks.com> Date: Thursday, August 1, 2019 at 3:07 AM To: Marco Gaido <marcogaid...@gmail.com> Cc: Wenchen Fan <cloud0...@gmail.com>, Hyukjin Kwon <gurwls...@gmail.com>, Russell Spitzer <russell.spit...@gmail.com>, Ryan Blue <rb...@netflix.com>, Reynold Xin <r...@databricks.com>, Matt Cheah <mch...@palantir.com>, Takeshi Yamamuro <linguin....@gmail.com>, Spark dev list <dev@spark.apache.org> Subject: Re: [Discuss] Follow ANSI SQL on table insertion Hi all, Let me explain a little bit on the proposal. By default, we follow the store assignment rules in table insertion. On invalid casting, the result is null. It's better than the behavior in Spark 2.x while keeping backward-compatibility. It is If users can't torrent the silently corrupting, they can enable the new mode which throws runtime exceptions. The proposal itself is quite complete. It satisfies different users to some degree. It is hard to avoid null in data processing anyway. For example, > select 2147483647 + 1 2147483647 is the max value of Int. And the result data type of pulsing two integers are supposed to be Integer type. Since the value of (2147483647 + 1) can't fit into Int, I think Spark return null or throw runtime exceptions in such case. (Someone can argue that we can always convert the result as wider types, but that's another topic about performance and DBMS behaviors) So, give a table t with an Int column, checking data type with Up-Cast can't avoid possible null values in the following SQL, as the result data type of (int_column_a + int_column_b) is int type. > insert into t select int_column_a + int_column_b from tbl_a, tbl_b; Furthermore, if Spark uses Up-Cast and a user's existing ETL job failed because of that, what should he/she do then? I think he/she will try adding "cast" to queries first. Maybe a project for unifying data schema over all data sources has to be done later on if he/she has enough resource. The upgrade can be painful because of the strict rules of Up-Cast, while the user scenario might be able to tolerate converting Double to Decimal, or Timestamp to Date. Gengliang On Thu, Aug 1, 2019 at 4:55 PM Marco Gaido <marcogaid...@gmail.com> wrote: Hi all, I agree that having both modes and let the user choose the one he/she wants is the best option (I don't see big arguments on this honestly). Once we have this, I don't see big differences on what is the default. What - I think - we still have to work on, is to go ahead with the "strict mode" work and provide a more convenient way for users to switch among the 2 options. I mean: currently we have one flag for throwing exception on overflow for operations on decimals, one for doing the same for operations on other data types and probably going ahead we will have more. I think in the end we will need to collect them all under an "umbrella" flag which lets the user simply switch between strict and non-strict mode. I also think that we will need to document this very well and give it particular attention in our docs, maybe with a dedicated section, in order to provide enough visibility on it to end users. Thanks, Marco Il giorno gio 1 ago 2019 alle ore 09:42 Wenchen Fan <cloud0...@gmail.com> ha scritto: Hi Hyukjin, I think no one here is against the SQL standard behavior, which is no corrupted data + runtime exception. IIUC the main argument here is: shall we still keep the existing "return null for invalid operations" behavior as default? Traditional RDBMS is usually used as the final destination of CLEAN data. It's understandable that they need high data quality and they try their best to avoid corrupted data at any cost. However, Spark is different. AFAIK Spark is usually used as an ETL tool, which needs to deal with DIRTY data. It's convenient if Spark returns null for invalid data and then I can filter them out later. I agree that "null" doesn't always mean invalid data, but it usually does. The "return null" behavior is there for many years and I don't see many people complain about it in the past. Recently there are several new projects at the storage side, which can host CLEAN data and can connect to Spark. This does raise a new requirement in Spark to improve data quality. The community is already working on it: now arithmetic operation fails if overflow happens(need to set a config). We are going to apply the same to cast as well, and any other expressions that return null for invalid input data. Rome was not built in a day. I think we need to keep the legacy "return null" behavior as default for a while, until we have enough confidence about the new ANSI mode. The default behavior should fit the majority of the users. I believe currently ETL is still the main use case of Spark and the "return null" behavior is useful to a lot of users. On Thu, Aug 1, 2019 at 8:55 AM Hyukjin Kwon <gurwls...@gmail.com> wrote: I am sorry I am asking a question without reading whole discussion after I replied. But why does Spark specifically needs to to it differently while ANCI standard, other DBMSes and other systems do? If there isn't a specific issue to Spark, that basically says they are all wrong. 2019년 8월 1일 (목) 오전 9:31, Russell Spitzer <russell.spit...@gmail.com>님이 작성: Another solution along those lines that I know we implemented for limited precision types is just to do a loud warning whenever you do such a cast. IE: Warning we are casting X to Y this may result in data loss. On Wed, Jul 31, 2019 at 7:08 PM Russell Spitzer <russell.spit...@gmail.com> wrote: I would argue "null" doesn't have to mean invalid. It could mean missing or deleted record. There is a lot of difference between missing record and invalid record. I definitely have no problem with two modes, but I think setting a parameter to enable lossy conversions is a fine tradeoff to avoid data loss for others. The impact then for those who don't care about lossy casting is an analysis level message "Types don't match, to enable lossy casting set some parameter" while the impact in the other direction is possibly invisible until it hits something critical downstream. On Wed, Jul 31, 2019 at 6:50 PM Ryan Blue <rb...@netflix.com> wrote: > you guys seem to be arguing no those users don't know what they are doing and > they should not exist. I'm not arguing that they don't exist. Just that the disproportionate impact of awareness about this behavior is much worse for people that don't know about it. And there are a lot of those people as well. On Wed, Jul 31, 2019 at 4:48 PM Ryan Blue <rb...@netflix.com> wrote: > "between a runtime error and an analysis-time error" → I think one of those > should be the default. If you're saying that the default should be an error of some kind, then I think we agree. I'm also fine with having a mode that allows turning off the error and silently replacing values with NULL... as long as it isn't the default and I can set the default for my platform to an analysis-time error. On Wed, Jul 31, 2019 at 4:42 PM Russell Spitzer <russell.spit...@gmail.com> wrote: I definitely view it as silently corrupting. If i'm copying over a dataset where some elements are null and some have values, how do I differentiate between my expected nulls and those that were added in silently in the cast? On Wed, Jul 31, 2019 at 6:15 PM Reynold Xin <r...@databricks.com> wrote: "between a runtime error and an analysis-time error" → I think one of those should be the default. Maybe we are talking past each other or I wasn't explaining clearly, but I don't think you understand what I said and the use cases out there. I as an end user could very well be fully aware of the consequences of exceptional values but I can choose to ignore them. This is especially common for data scientists who are doing some quick and dirty analysis or exploration. You can't deny this large class of use cases out there (probably makes up half of Spark use cases actually). Also writing out the exceptional cases as null are not silently corrupting them. The engine is sending an explicit signal that the value is no longer valid given the constraint. Not sure if this is the best analogy, but think about checked exceptions in Java. It's great for writing low level code in which error handling is paramount, e.g. storage systems, network layers. But in most high level applications people just write boilerplate catches that are no-ops, because they have other priorities and they can tolerate mishandling of exceptions, although often maybe they shouldn't. On Wed, Jul 31, 2019 at 2:55 PM, Ryan Blue <rb...@netflix.com> wrote: Another important aspect of this problem is whether a user is conscious of the cast that is inserted by Spark. Most of the time, users are not aware of casts that are implicitly inserted, and that means replacing values with NULL would be a very surprising behavior. The impact of this choice affects users disproportionately: someone that knows about inserted casts is mildly annoyed when required to add an explicit cast, but a user that doesn't know an inserted cast is dropping values is very negatively impacted and may not discover the problem until it is too late. That disproportionate impact is what makes me think that it is not okay for Spark to silently replace values with NULL, even if that's what ANSI would allow. Other databases also have the ability to reject null values in tables, providing extra insurance against the problem, but Spark doesn't have required columns in its DDL. So while I agree with Reynold that there is a trade-off, I think that trade-off makes the choice between a runtime error and an analysis-time error. I'm okay with either a runtime error as the default or an analysis error as the default, as long as there is a setting that allows me to choose one for my deployment. On Wed, Jul 31, 2019 at 10:39 AM Reynold Xin <r...@databricks.com> wrote: OK to push back: "disagreeing with the premise that we can afford to not be maximal on standard 3. The correctness of the data is non-negotiable, and whatever solution we settle on cannot silently adjust the user’s data under any circumstances." This blanket statement sounds great on surface, but there are a lot of subtleties. "Correctness" is absolutely important, but engineering/prod development are often about tradeoffs, and the industry has consistently traded correctness for performance or convenience, e.g. overflow checks, null pointers, consistency in databases ... It all depends on the use cases and to what degree use cases can tolerate. For example, while I want my data engineering production pipeline to throw any error when the data doesn't match my expectations (e.g. type widening, overflow), if I'm doing some quick and dirty data science, I don't want the job to just fail due to outliers. On Wed, Jul 31, 2019 at 10:13 AM, Matt Cheah <mch...@palantir.com> wrote: Sorry I meant the current behavior for V2, which fails the query compilation if the cast is not safe. Agreed that a separate discussion about overflow might be warranted. I’m surprised we don’t throw an error now, but it might be warranted to do so. -Matt Cheah From: Reynold Xin <r...@databricks.com> Date: Wednesday, July 31, 2019 at 9:58 AM To: Matt Cheah <mch...@palantir.com> Cc: Russell Spitzer <russell.spit...@gmail.com>, Takeshi Yamamuro <linguin....@gmail.com>, Gengliang Wang <gengliang.w...@databricks.com>, Ryan Blue <rb...@netflix.com>, Spark dev list <dev@spark.apache.org>, Hyukjin Kwon <gurwls...@gmail.com>, Wenchen Fan <cloud0...@gmail.com> Subject: Re: [Discuss] Follow ANSI SQL on table insertion Error! Filename not specified. Matt what do you mean by maximizing 3, while allowing not throwing errors when any operations overflow? Those two seem contradicting. On Wed, Jul 31, 2019 at 9:55 AM, Matt Cheah <mch...@palantir.com> wrote: I’m -1, simply from disagreeing with the premise that we can afford to not be maximal on standard 3. The correctness of the data is non-negotiable, and whatever solution we settle on cannot silently adjust the user’s data under any circumstances. I think the existing behavior is fine, or perhaps the behavior can be flagged by the destination writer at write time. -Matt Cheah From: Hyukjin Kwon <gurwls...@gmail.com> Date: Monday, July 29, 2019 at 11:33 PM To: Wenchen Fan <cloud0...@gmail.com> Cc: Russell Spitzer <russell.spit...@gmail.com>, Takeshi Yamamuro <linguin....@gmail.com>, Gengliang Wang <gengliang.w...@databricks.com>, Ryan Blue <rb...@netflix.com>, Spark dev list <dev@spark.apache.org> Subject: Re: [Discuss] Follow ANSI SQL on table insertion >From my look, +1 on the proposal, considering ASCI and other DBMSes in general. 2019년 7월 30일 (화) 오후 3:21, Wenchen Fan <cloud0...@gmail.com>님이 작성: We can add a config for a certain behavior if it makes sense, but the most important thing we want to reach an agreement here is: what should be the default behavior? Let's explore the solution space of table insertion behavior first: At compile time, 1. always add cast 2. add cast following the ASNI SQL store assignment rule (e.g. string to int is forbidden but long to int is allowed) 3. only add cast if it's 100% safe At runtime, 1. return null for invalid operations 2. throw exceptions at runtime for invalid operations The standards to evaluate a solution: 1. How robust the query execution is. For example, users usually don't want to see the query fails midway. 2. how tolerant to user queries. For example, a user would like to write long values to an int column as he knows all the long values won't exceed int range. 3. How clean the result is. For example, users usually don't want to see silently corrupted data (null values). The current Spark behavior for Data Source V1 tables: always add cast and return null for invalid operations. This maximizes standard 1 and 2, but the result is least clean and users are very likely to see silently corrupted data (null values). The current Spark behavior for Data Source V2 tables (new in Spark 3.0): only add cast if it's 100% safe. This maximizes standard 1 and 3, but many queries may fail to compile, even if these queries can run on other SQL systems. Note that, people can still see silently corrupted data because cast is not the only one that can return corrupted data. Simple operations like ADD can also return corrected data if overflow happens. e.g. INSERT INTO t1 (intCol) SELECT anotherIntCol + 100 FROM t2 The proposal here: add cast following ANSI SQL store assignment rule, and return null for invalid operations. This maximizes standard 1, and also fits standard 2 well: if a query can't compile in Spark, it usually can't compile in other mainstream databases as well. I think that's tolerant enough. For standard 3, this proposal doesn't maximize it but can avoid many invalid operations already. Technically we can't make the result 100% clean at compile-time, we have to handle things like overflow at runtime. I think the new proposal makes more sense as the default behavior. On Mon, Jul 29, 2019 at 8:31 PM Russell Spitzer <russell.spit...@gmail.com> wrote: I understand spark is making the decisions, i'm say the actual final effect of the null decision would be different depending on the insertion target if the target has different behaviors for null. On Mon, Jul 29, 2019 at 5:26 AM Wenchen Fan <cloud0...@gmail.com> wrote: > I'm a big -1 on null values for invalid casts. This is why we want to introduce the ANSI mode, so that invalid cast fails at runtime. But we have to keep the null behavior for a while, to keep backward compatibility. Spark returns null for invalid cast since the first day of Spark SQL, we can't just change it without a way to restore to the old behavior. I'm OK with adding a strict mode for the upcast behavior in table insertion, but I don't agree with making it the default. The default behavior should be either the ANSI SQL behavior or the legacy Spark behavior. > other modes should be allowed only with strict warning the behavior will be > determined by the underlying sink. Seems there is some misunderstanding. The table insertion behavior is fully controlled by Spark. Spark decides when to add cast and Spark decided whether invalid cast should return null or fail. The sink is only responsible for writing data, not the type coercion/cast stuff. On Sun, Jul 28, 2019 at 12:24 AM Russell Spitzer <russell.spit...@gmail.com> wrote: I'm a big -1 on null values for invalid casts. This can lead to a lot of even more unexpected errors and runtime behavior since null is 1. Not allowed in all schemas (Leading to a runtime error anyway) 2. Is the same as delete in some systems (leading to data loss) And this would be dependent on the sink being used. Spark won't just be interacting with ANSI compliant sinks so I think it makes much more sense to be strict. I think Upcast mode is a sensible default and other modes should be allowed only with strict warning the behavior will be determined by the underlying sink. On Sat, Jul 27, 2019 at 8:05 AM Takeshi Yamamuro <linguin....@gmail.com> wrote: Hi, all +1 for implementing this new store cast mode. >From a viewpoint of DBMS users, this cast is pretty common for INSERTs and I >think this functionality could promote migrations from existing DBMSs to Spark. The most important thing for DBMS users is that they could optionally choose this mode when inserting data. Therefore, I think it might be okay that the two modes (the current upcast mode and the proposed store cast mode) co-exist for INSERTs. (There is a room to discuss which mode is enabled by default though...) IMHO we'll provide three behaviours below for INSERTs; - upcast mode - ANSI store cast mode and runtime exceptions thrown for invalid values - ANSI store cast mode and null filled for invalid values On Sat, Jul 27, 2019 at 8:03 PM Gengliang Wang <gengliang.w...@databricks.com> wrote: Hi Ryan, Thanks for the suggestions on the proposal and doc. Currently, there is no data type validation in table insertion of V1. We are on the same page that we should improve it. But using UpCast is from one extreme to another. It is possible that many queries are broken after upgrading to Spark 3.0. The rules of UpCast are too strict. E.g. it doesn't allow assigning Timestamp type to Date Type, as there will be "precision loss". To me, the type coercion is reasonable and the "precision loss" is under expectation. This is very common in other SQL engines. As long as Spark is following the ANSI SQL store assignment rules, it is users' responsibility to take good care of the type coercion in data writing. I think it's the right decision. > But the new behavior is only applied in DataSourceV2, so it won’t affect > existing jobs until sources move to v2 and break other behavior anyway. Eventually, most sources are supposed to be migrated to DataSourceV2 V2. I think we can discuss and make a decision now. > Fixing the silent corruption by adding a runtime exception is not a good > option, either. The new optional mode proposed in https://issues.apache.org/jira/browse/SPARK-28512 [issues.apache.org] is disabled by default. This should be fine. On Sat, Jul 27, 2019 at 10:23 AM Wenchen Fan <cloud0...@gmail.com> wrote: I don't agree with handling literal values specially. Although Postgres does it, I can't find anything about it in the SQL standard. And it introduces inconsistent behaviors which may be strange to users: * What about something like "INSERT INTO t SELECT float_col + 1.1"? * The same insert with a decimal column as input will fail even when a decimal literal would succeed * Similar insert queries with "literal" inputs can be constructed through layers of indirection via views, inline views, CTEs, unions, etc. Would those decimals be treated as columns and fail or would we attempt to make them succeed as well? Would users find this behavior surprising? Silently corrupt data is bad, but this is the decision we made at the beginning when design Spark behaviors. Whenever an error occurs, Spark attempts to return null instead of runtime exception. Recently we provide configs to make Spark fail at runtime for overflow, but that's another story. Silently corrupt data is bad, runtime exception is bad, and forbidding all the table insertions that may fail(even with very little possibility) is also bad. We have to make trade-offs. The trade-offs we made in this proposal are: * forbid table insertions that are very like to fail, at compile time. (things like writing string values to int column) * allow table insertions that are not that likely to fail. If the data is wrong, don't fail, insert null. * provide a config to fail the insertion at runtime if the data is wrong. > But the new behavior is only applied in DataSourceV2, so it won’t affect > existing jobs until sources move to v2 and break other behavior anyway. When users write SQL queries, they don't care if a table is backed by Data Source V1 or V2. We should make sure the table insertion behavior is consistent and reasonable. Furthermore, users may even not care if the SQL queries are run in Spark or other RDBMS, it's better to follow SQL standard instead of introducing a Spark-specific behavior. We are not talking about a small use case like allowing writing decimal literal to float column, we are talking about a big goal to make Spark compliant to SQL standard, w.r.t. https://issues.apache.org/jira/browse/SPARK-26217 [issues.apache.org] . This proposal is a sub-task of it, to make the table insertion behavior follow SQL standard. On Sat, Jul 27, 2019 at 1:35 AM Ryan Blue <rb...@netflix.com> wrote: I don’t think this is a good idea. Following the ANSI standard is usually fine, but here it would silently corrupt data. >From your proposal doc, ANSI allows implicitly casting from long to int (any >numeric type to any other numeric type) and inserts NULL when a value >overflows. That would drop data values and is not safe. Fixing the silent corruption by adding a runtime exception is not a good option, either. That puts off the problem until much of the job has completed, instead of catching the error at analysis time. It is better to catch this earlier during analysis than to run most of a job and then fail. In addition, part of the justification for using the ANSI standard is to avoid breaking existing jobs. But the new behavior is only applied in DataSourceV2, so it won’t affect existing jobs until sources move to v2 and break other behavior anyway. I think that the correct solution is to go with the existing validation rules that require explicit casts to truncate values. That still leaves the use case that motivated this proposal, which is that floating point literals are parsed as decimals and fail simple insert statements. We already came up with two alternatives to fix that problem in the DSv2 sync and I think it is a better idea to go with one of those instead of “fixing” Spark in a way that will corrupt data or cause runtime failures. On Thu, Jul 25, 2019 at 9:11 AM Wenchen Fan <cloud0...@gmail.com> wrote: I have heard about many complaints about the old table insertion behavior. Blindly casting everything will leak the user mistake to a late stage of the data pipeline, and make it very hard to debug. When a user writes string values to an int column, it's probably a mistake and the columns are misordered in the INSERT statement. We should fail the query earlier and ask users to fix the mistake. In the meanwhile, I agree that the new table insertion behavior we introduced for Data Source V2 is too strict. It may fail valid queries unexpectedly. In general, I support the direction of following the ANSI SQL standard. But I'd like to do it with 2 steps: 1. only add cast when the assignment rule is satisfied. This should be the default behavior and we should provide a legacy config to restore to the old behavior. 2. fail the cast operation at runtime if overflow happens. AFAIK Marco Gaido is working on it already. This will have a config as well and by default we still return null. After doing this, the default behavior will be slightly different from the SQL standard (cast can return null), and users can turn on the ANSI mode to fully follow the SQL standard. This is much better than before and should prevent a lot of user mistakes. It's also a reasonable choice to me to not throw exceptions at runtime by default, as it's usually bad for long-running jobs. Thanks, Wenchen On Thu, Jul 25, 2019 at 11:37 PM Gengliang Wang <gengliang.w...@databricks.com> wrote: Hi everyone, I would like to discuss the table insertion behavior of Spark. In the current data source V2, only UpCast is allowed for table insertion. I think following ANSI SQL is a better idea. For more information, please read the Discuss: Follow ANSI SQL on table insertion [docs.google.com] Please let me know if you have any thoughts on this. Regards, Gengliang -- Ryan Blue Software Engineer Netflix -- --- Takeshi Yamamuro -- Ryan Blue Software Engineer Netflix -- Ryan Blue Software Engineer Netflix -- Ryan Blue Software Engineer Netflix
smime.p7s
Description: S/MIME cryptographic signature
