16 KiB
16 KiB
type:: REVIEWS tags:: year:: 2024 venue:: MODELS full-title:: The Architecture of Senderismo: A Framework for an Automated Schema and Data Migration between different Data Stores date-start:: 23-04-2024 - 15:57 date-submitted:: 01-05-2024 external-links:: Add Review on MODELS'24 Submission 8050 (easychair.org) status:: DONE deadline-submission:: 23-04-2024 file:: @MODELS24_paper_8050.pdf parent:: todoist:: https://app.todoist.com/app/task/8050-the-architecture-of-senderismo-a-framework-for-an-automated-schema-and-data-7858701449
- ### [[Highlights]]
collapsed:: true
-
- # Annotazioni
- (1/5/2024, 22:11:32)
- - “challenges” ([“MODELS24_paper_8050.pdf”, p. 1](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=1&annotation=BANDSIYS)) #f0ff00
- *These should be described better. What challenges in particular?*
- - “an” ([“MODELS24_paper_8050.pdf”, p. 1](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=1&annotation=XLCCI68L)) #ff6666
- *a*
- - “an” ([“MODELS24_paper_8050.pdf”, p. 1](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=1&annotation=F5MATFR3)) #f0ff00
- *A*
- - “automatic data and schema migration between all kinds of data stores, including necessary optimizations of the schema.” ([“MODELS24_paper_8050.pdf”, p. 1](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=1&annotation=VVAPJSS6)) #5fb236
- - “all kinds of data stores,” ([“MODELS24_paper_8050.pdf”, p. 1](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=1&annotation=ES4RNL7Z)) #f0ff00
- *All kinds?*
- - “optimizations” ([“MODELS24_paper_8050.pdf”, p. 1](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=1&annotation=QRWKUJSI)) #f0ff00
- *What kinds of optimization?*
- - “suitable architecture represents an important basis regarding an implementation.” ([“MODELS24_paper_8050.pdf”, p. 1](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=1&annotation=REHXLSWU)) #ffd400
- *suitable architecture for what?*
- - “The biggest challenges are how to handle the complexity of such a system, the required flexibility in terms of extensions and interchangeability of the individual "components", as well as the scalability to process large amounts of data.” ([“MODELS24_paper_8050.pdf”, p. 1](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=1&annotation=SU4BY2GK)) #f0ff00
- *Very vague so far.*
- - “First approaches to a flexible and scalable architecture based on microservices” ([“MODELS24_paper_8050.pdf”, p. 1](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=1&annotation=I3JVGZUK)) #ffd400
- *Does not parse*
- - “First” ([“MODELS24_paper_8050.pdf”, p. 1](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=1&annotation=VPD6RGZT)) #f0ff00
- *First?*
- - “architecture to satisfy the previously established criteria is discussed.” ([“MODELS24_paper_8050.pdf”, p. 1](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=1&annotation=6PCIKSWR)) #f0ff00
- - “we using a subset of” ([“MODELS24_paper_8050.pdf”, p. 2](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=2&annotation=U7WMT8SE)) #ff4400
- *Many grammatical errors.*
- - “AUTOMATED SCHEMA AND DATA MIGRATION” ([“MODELS24_paper_8050.pdf”, p. 2](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=2&annotation=WGTUUHXE)) #f0ff00
- *Check if typhon and alike are considered as starting point.*
- - “In addition to our own implementation of a schema extraction algorithm with focus on reference detection and support for all types of data stores (including polystores), we have integrated the graph extraction of [16] and the implementation of [8], which support various aggregator-oriented data stores.” ([“MODELS24_paper_8050.pdf”, p. 2](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=2&annotation=QXSWALC4)) #ffd400
- *This means that you have developed your own schema extraction approach by integrating [16] and [8]. As it is the section, is not effective in distilling the peculiarities of what you have done. I suggest to revise the section to directly presenting what you have done and in what respect to improve existing techniques/approaches. Currently, the section contains too many references and it is not self-container by compromising the readability of the whole section.
- The mentioned polystore concept is not supported by the technical discussion. Managing polystores is very difficult and currently the paper is not exploring and showing how the proposed approach is able to deal with polystores.*
- - “For the schema management and data migration in our approach, we use the M-model from [9]. In addition, Darwin’s schema evolution language has been extended to include further schema modification operations (SMO)s, which are required in our migration and optimization process.” ([“MODELS24_paper_8050.pdf”, p. 2](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=2&annotation=956YZCIX)) #ffd400
- *See my previous comment. It applies also to this section.*
- - “continious example.” ([“MODELS24_paper_8050.pdf”, p. 3](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=3&annotation=4ZYYW3Y7)) #ff4400
- *Continuous*
- - “target-specific meta model” ([“MODELS24_paper_8050.pdf”, p. 3](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=3&annotation=8X4588F2)) #ffd400
- *At this stage, it is not clear what is the target-specific metamodel. Moreover, Fig. 5 and FIg, 6 should be represented in terms of object diagrams showing the instances of metaclasses of the source and target metamodels (which unfortunately have not shown in the paper).*
- - “Figure 6 shows the target-specific meta model after applying the SMOs and Figure 7 illustrates what the physical model of Neo4j would look like at this point.” ([“MODELS24_paper_8050.pdf”, p. 3](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=3&annotation=LN8RKWGY)) #ffd400
- *The implemented operations need to be explained. Currently, the reaser is supposed to read Listing 1 and draw her considerations/conclusions on her own.*
- - “.3 Data & Workload Analysis, Schema Optimization” ([“MODELS24_paper_8050.pdf”, p. 3](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=3&annotation=CKHQEVW2)) #ffd400
- *The whole section, and the second half in particular, is very verbose and does not clearly presents what are the novelties and the contributions of the authors concerning the work done for data and workload analysis and the consequent schema optimization. I suggest the authors to clearly state early in the paper what are the main technical contributions that goes beyond the usage and application of existing techniques to implement the workflow shown in Fig. 4.*
- - “Listing 1: Schema modification operations of the continious example,” ([“MODELS24_paper_8050.pdf”, p. 3](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=3&annotation=T7CXFS7V)) #f0ff00
- *Who writ this?*
- - “generated by the automated schema optimization, a” ([“MODELS24_paper_8050.pdf”, p. 5](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=5&annotation=X6GWNX6C)) #ffd400
- *Who performs the generation of the SMOs?*
- - “Listing 3 shows the SMOs of these manual optimizations,” ([“MODELS24_paper_8050.pdf”, p. 5](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=5&annotation=9F9XLMWJ)) #ffd400
- *This sentence seems to be contracting the previous one. It is not clear if the SMOs shown in Listing 3 are (fully/partially) automatically generated or manually specified. Moreover, a detailed presentation of the logic underpinning the performed optimization is required.*
- - “operstions” ([“MODELS24_paper_8050.pdf”, p. 5](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=5&annotation=FFBRIE9U)) #ff6666
- - “aplying” ([“MODELS24_paper_8050.pdf”, p. 5](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=5&annotation=Q9FXRF3M)) #ff6666
- - “operstions” ([“MODELS24_paper_8050.pdf”, p. 5](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=5&annotation=27EZWAMJ)) #ff6666
- - “aplying” ([“MODELS24_paper_8050.pdf”, p. 6](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=6&annotation=BZVYHPWF)) #ff6666
- - “apying” ([“MODELS24_paper_8050.pdf”, p. 6](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=6&annotation=5B26UUUT)) #ff6666
- - “apying” ([“MODELS24_paper_8050.pdf”, p. 6](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=6&annotation=6JCEJPFW)) #ff6666
- - “The migration of data primarily concerns the mapping of the data structures of the source store to the data structures of the target store. With the use of the meta model (M-model), the schema changes can be easily applied during the data migration. Figure 4 illustrates the process of data migration in detail.” ([“MODELS24_paper_8050.pdf”, p. 6](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=6&annotation=8IDTJ29I)) #ffd400
- *The paper contains several sentences like this one that vaguously present a concept without providing technical and convincing arguments supporting it.*
- - “We developing” ([“MODELS24_paper_8050.pdf”, p. 7](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=7&annotation=EVB2KDMI)) #ff6666
- - “a generic approach based on Apache Spark2 that on the one hand enables scalable data migration between different relational and NoSQL stores (in all directions) and on the other hand can handle schema changes (optimizations) during data migration. The use of Apache Spark also has the advantage that a large number of connectors for relational and NoSQL database systems are already available.” ([“MODELS24_paper_8050.pdf”, p. 7](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=7&annotation=KRA6RDAD)) #ffd400
- *The related work part of this section can be shortened to give more details about what the authors have done as novel contribution concerning the schema and data migration task.*
- - “The choice of a suitable design is an important prerequisite for the implementation. Since automatic optimization and migration between different data stores is a complex task consisting of different components (see Figure 4), for which various research prototypes are available, one goal is to easily integrate existing approaches in such a system.” ([“MODELS24_paper_8050.pdf”, p. 7](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=7&annotation=L53WSQ4R)) #ffd400
- *It seems that integration is main contribution of this paper, isn't it?*
- - “all types of data stores including automated schema optimizations.” ([“MODELS24_paper_8050.pdf”, p. 7](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=7&annotation=5N58PWJP)) #ffd400
- *As mentioned by the authors while overviewing the different related work, there are many approaches addressing the problem of the authors. It is not clear, what are the novelties of the proposed approach.*
- - “highly flexible and scalable design for an automated schema and data migration process between all types of data stores including schema optimizations.” ([“MODELS24_paper_8050.pdf”, p. 8](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=8&annotation=MTLYSUFN)) #ffd400
- *No quantitative and qualitative analysis supporting sentences like this one!*
- - “The Currently implemented components are: Migration Manager (includes core logic), Web GUI, various Schema Extraction Implementation, Schema Manager (currently without the use the extended schema modification language), our Schema Optimizer approach and Data Migration. Furthermore, a first prototype of the Schema Editor has been implemented.” ([“MODELS24_paper_8050.pdf”, p. 8](zotero://select/library/items/DN3BQIXA)) ([pdf](zotero://open-pdf/library/items/GGMG22IY?page=8&annotation=QS5G6RDC)) #ffd400
- *Such clarifications should not come to the conclusion section. They should be presented earlier in the paper.*
- ### [[Comments]]
- #.tabular
- Summary of the paper
- The paper introduces Senderismo, a comprehensive framework designed to facilitate schema and data migration tasks across various data stores. It consists of a collection of microservices, each responsible for specific functionalities within the migration process. The primary objective of Senderismo is to automate the complex and labor-intensive task of schema and data migration, which is often necessary when transitioning between different data storage technologies or optimizing existing data schemas.
- _Summary of main points for/against the paper_
- + Relevant problem
- - Lack of Clarity
- - Limited evaluation
- - Paper organization to be improved with technical details
- - Insufficient explanation of the novel contributions and improvements over existing approaches
- Detailed evaluation
- While the topic is relevant and the paper attempts to address significant challenges, there are several areas where clarity and depth of discussion could be improved as discussed below:
- The paper mentions various challenges related to handling complexity, flexibility, and scalability related to schema and data migration but lacks specific details on these challenges. For instance, what aspects of complexity are being addressed? What kinds of optimizations are performed? I suggest improving the paper by providing clearer descriptions of challenges, optimizations, and architectural choices that permit overcoming existing techniques' limitations. As mentioned by the authors, while overviewing the different related work, there are many approaches addressing the problems of data and schema migration between different data stores. However, it is not clear what the novelties of the approach proposed in this paper are. Is it an integration effort to assemble various components already presented or available elsewhere?
- The integration of various schema extraction techniques is mentioned, but the paper lacks a clear presentation of the novel contributions in this regard. How does the proposed approach improve upon existing techniques? Provide detailed explanations of novel algorithms, methodologies, or optimizations.
- There is confusion regarding the definition of the mentioned SMOs. Are they automatically generated or manually specified? A detailed explanation of the logic behind the optimizations is needed.
- The process of data migration is vaguely presented without providing convincing technical arguments. What are the mechanisms underlying the migration process?
- In the conclusion section, the paper briefly mentions implemented components but fails to provide detailed explanations. These clarifications should be presented earlier in the paper for better context. Explain how various components are integrated and provide insights into the evaluation methodology. Include quantitative and qualitative analyses to support claims of flexibility and scalability.
-
- _Questions to authors (rebuttal phase)_
- N/A
- ### [[REVIEWS/Notes]]