Uploaded image for project: 'Couchbase Server'
  1. Couchbase Server
  2. MB-7852

[Doc'd] Complete productization of Replica Read API

    Details

    • Type: Task
    • Status: Closed
    • Priority: Minor
    • Resolution: Fixed
    • Affects Version/s: 2.0
    • Fix Version/s: 2.1.0
    • Component/s: None
    • Security Level: Public
    • Flagged:
      Release Note

      Description

      Replica Read API implementation is ready but testing and related documentation have not completed up to the product-feature ready state.

      No reviews matched the request. Check your Options in the drop-down menu of this sections header.

        Activity

        Hide
        perry Perry Krug added a comment - - edited

        Thanks Karen, looks great. Some quick comments from me:
        [FIXED]-Page 128 under "Replica Read": "we have binary protocol" should be "we have a binary protocol"
        [FIXED]-Page 128: "longer to for human intervention" should be "longer for human intervention"
        [MOVED HIGHER UP]-Page 128: I think the link to more information on failing over nodes should be higher up in the paragraph (after the 3rd paragram) and/or separated from the discussion about the decision to use the replica read
        [FIXED]-Page 128: This seems a little harsh: "We do not recommend you use replica read for any scenario, especially if you do not care about a very high level of availability". I would reword it a bit along the lines of "Using the replica read functionality can introduce the possibility of inconsistent data and our general recommendation is to have application-level logic to deal with short periods of unavailability" CHANGED TO: If you use replica read, it will introduce the possibility that a client gets inconsistent data from the cluster; for this reason we generally recommend you have your application logic handle shorts periods of unavailability.

        Show
        perry Perry Krug added a comment - - edited Thanks Karen, looks great. Some quick comments from me: [FIXED] -Page 128 under "Replica Read": "we have binary protocol" should be "we have a binary protocol" [FIXED] -Page 128: "longer to for human intervention" should be "longer for human intervention" [MOVED HIGHER UP] -Page 128: I think the link to more information on failing over nodes should be higher up in the paragraph (after the 3rd paragram) and/or separated from the discussion about the decision to use the replica read [FIXED] -Page 128: This seems a little harsh: "We do not recommend you use replica read for any scenario, especially if you do not care about a very high level of availability". I would reword it a bit along the lines of "Using the replica read functionality can introduce the possibility of inconsistent data and our general recommendation is to have application-level logic to deal with short periods of unavailability" CHANGED TO: If you use replica read, it will introduce the possibility that a client gets inconsistent data from the cluster; for this reason we generally recommend you have your application logic handle shorts periods of unavailability.
        Hide
        kzeller kzeller added a comment -

        Hi Deep,

        You should talk to Jin for more clarification on code snippets, but my understanding is that the binary we show IS everything you send for the request. The only parameter you alter is the key you send in binary.

        On the second point, we made an explicit decision for 1.8 and 2.0 that we NOT document the binary as the other APIs because it is for people building client libraries, not for people who are using a language-based client library to build a web application. It is for a separate category of developers who are building, for example a Erlang driver or NodeJs driver for Couchbase.

        Regards,

        Karen

        From: Deepkaran Salooja <deepkaran.salooja@globallogic.com>
        Date: Wednesday, May 22, 2013 5:31 AM
        To: Karen Zeller <karen.zeller@couchbase.com>
        Cc: Iryna <irynamironava@gmail.com>, Maria McDuff <Maria@couchbase.com>, Iryna Mironava <iryna@couchbase.com>
        Subject: Re: [DOCS Draft Review: Replica Read Binary]

        Hi Karen,

        The explanation looks good. Few comments:

        • We can have the name/details(e.g. how many arguments) of the API for replica read or some code snippet like we have in case of CAS/append/prepend, which would clarify the usage.
        • Also I am a little unclear about keeping this under the section "Chapter8. Developing a Client Library" rather than along with other APIs we provide.

        Thanks,
        Deep

        Show
        kzeller kzeller added a comment - Hi Deep, You should talk to Jin for more clarification on code snippets, but my understanding is that the binary we show IS everything you send for the request. The only parameter you alter is the key you send in binary. On the second point, we made an explicit decision for 1.8 and 2.0 that we NOT document the binary as the other APIs because it is for people building client libraries, not for people who are using a language-based client library to build a web application. It is for a separate category of developers who are building, for example a Erlang driver or NodeJs driver for Couchbase. Regards, Karen From: Deepkaran Salooja <deepkaran.salooja@globallogic.com> Date: Wednesday, May 22, 2013 5:31 AM To: Karen Zeller <karen.zeller@couchbase.com> Cc: Iryna <irynamironava@gmail.com>, Maria McDuff <Maria@couchbase.com>, Iryna Mironava <iryna@couchbase.com> Subject: Re: [DOCS Draft Review: Replica Read Binary] Hi Karen, The explanation looks good. Few comments: We can have the name/details(e.g. how many arguments) of the API for replica read or some code snippet like we have in case of CAS/append/prepend, which would clarify the usage. Also I am a little unclear about keeping this under the section "Chapter8. Developing a Client Library" rather than along with other APIs we provide. Thanks, Deep
        Hide
        kzeller kzeller added a comment -

        Sent 5/20

        Attached is draft content for replica read as a binary protocol:

        "Replica Read", pages 128-129 of Developer Guide

        Mandatory Reviewers: Jin Lim, Irina, Deep

        Optional Reviewers: PerryK, TimS, FrankW

        Bug tracking doc update: http://www.couchbase.com/issues/browse/MB-7852
        Review timeframe: *Please provide your review input by EOD Friday, May 24 of this week.

        *Please add comments directly to the PDF or to the bug link above.

        Show
        kzeller kzeller added a comment - Sent 5/20 Attached is draft content for replica read as a binary protocol: "Replica Read", pages 128-129 of Developer Guide Mandatory Reviewers: Jin Lim, Irina, Deep Optional Reviewers: PerryK, TimS, FrankW Bug tracking doc update: http://www.couchbase.com/issues/browse/MB-7852 Review timeframe: *Please provide your review input by EOD Friday, May 24 of this week. *Please add comments directly to the PDF or to the bug link above.
        Hide
        kzeller kzeller added a comment -
        Show
        kzeller kzeller added a comment - Final drafted to be published: http://www.couchbase.com/docs/couchbase-devguide-2.0/cb-protocol-replica-read.html
        Hide
        kzeller kzeller added a comment -
        Show
        kzeller kzeller added a comment - Final drafted to be published: http://www.couchbase.com/docs/couchbase-devguide-2.0/cb-protocol-replica-read.html

          People

          • Assignee:
            kzeller kzeller
            Reporter:
            jin Jin Lim (Inactive)
          • Votes:
            0 Vote for this issue
            Watchers:
            6 Start watching this issue

            Dates

            • Created:
              Updated:
              Resolved:

              Gerrit Reviews

              There are no open Gerrit changes