Uploaded image for project: 'Couchbase Server'
  1. Couchbase Server
  2. MB-6122

Better error responses for append/prepend "not stored"

    Details

    • Type: Improvement
    • Status: Closed
    • Priority: Major
    • Resolution: Duplicate
    • Affects Version/s: 1.8.1
    • Fix Version/s: 3.0
    • Component/s: couchbase-bucket
    • Security Level: Public
    • Labels:
      None

      Description

      At the moment, the server only responds with "not stored" to append/prepend operations for multiple possible reasons:
      -The item was over 20mb
      -The item didn't exist
      -others?

      It would be very helpful to our users to have these split out with separate error messages. For instance, incr/decr respond with "not found" as opposed to "not stored" when the item doesn't exist, and set operations return with "value too large" when it is over 20mb.

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

        Activity

        Hide
        mikew Mike Wiederhold added a comment -

        We can add some better messages for given error codes, but at the end of the day it is up to the clients to display these and users to ask for them. I know capability is available in the java sdk, but I doubt it is in many of our other clients.

        Show
        mikew Mike Wiederhold added a comment - We can add some better messages for given error codes, but at the end of the day it is up to the clients to display these and users to ask for them. I know capability is available in the java sdk, but I doubt it is in many of our other clients.
        Hide
        ingenthr Matt Ingenthron added a comment -

        Actually, it is more likely in the scripting ones, such as PHP and Ruby. They're a bit more freeform. I'm not sure about .NET, but I've added John as a watcher here.

        John, can you comment?

        Show
        ingenthr Matt Ingenthron added a comment - Actually, it is more likely in the scripting ones, such as PHP and Ruby. They're a bit more freeform. I'm not sure about .NET, but I've added John as a watcher here. John, can you comment?
        Hide
        john John Zablocki (Inactive) added a comment -

        The .NET client is echoing back the status codes/messages, so if the server response changed, so should the client's handling of it.

        Show
        john John Zablocki (Inactive) added a comment - The .NET client is echoing back the status codes/messages, so if the server response changed, so should the client's handling of it.
        Hide
        mikew Mike Wiederhold added a comment -

        This stuff needs to be taken care of in memcached and that poses an issue because memcached receives an error code that could have multiple meanings.

        Show
        mikew Mike Wiederhold added a comment - This stuff needs to be taken care of in memcached and that poses an issue because memcached receives an error code that could have multiple meanings.
        Hide
        trond Trond Norbye added a comment -

        Mike: it is the error code returned from the engine that is returned back to the caller (it goes through the same codepath as set/add/replace/cas etc). Given that this method is implemented inside ep-engine it may return the appropriate error message back. Do you still want me to look at it?

        Show
        trond Trond Norbye added a comment - Mike: it is the error code returned from the engine that is returned back to the caller (it goes through the same codepath as set/add/replace/cas etc). Given that this method is implemented inside ep-engine it may return the appropriate error message back. Do you still want me to look at it?
        Hide
        cihan Cihan Biyikoglu added a comment -

        this seems like a dup of https://www.couchbase.com/issues/browse/MB-11332. I think we should close until we get a backward compat flag that can help implement such breaking changes.

        Show
        cihan Cihan Biyikoglu added a comment - this seems like a dup of https://www.couchbase.com/issues/browse/MB-11332 . I think we should close until we get a backward compat flag that can help implement such breaking changes.

          People

          • Assignee:
            mikew Mike Wiederhold
            Reporter:
            perry Perry Krug
          • Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

            Dates

            • Created:
              Updated:
              Resolved:

              Gerrit Reviews

              There are no open Gerrit changes