Francesco Mari updated OAK-2498:
    Attachment: OAK-2498-01.patch

OAK-2498-01.patch is a first take at this problem. I implemented every type 
discussed so far, except for the super root. Figuring out when and where to use 
the super root type is not a trivial issue.

Implementing the additional types as done in OAK-2498-01.patch still doesn't 
make the segment easily parseable. The main reason is that the bucket record 
type is not parseable without knowing the size of the list upfront. Given a 
pointer to a bucket record, it is impossible to know where the content of the 
bucket record actually ends. The same issue might occur for records of type 

We would still have some advantages by applying this patch. Since the types are 
more explicit, we can extract more detailed usage information - e.g. the exact 
type of lists used throughout the segment store.

I will start with a partial implementation of this patch focused on the type of 
binary IDs. This is necessary for OAK-4740 and for the correct functioning of 
the cold standby client.

> Root record references provide too little context for parsing a segment
> -----------------------------------------------------------------------
>                 Key: OAK-2498
>                 URL: https://issues.apache.org/jira/browse/OAK-2498
>             Project: Jackrabbit Oak
>          Issue Type: Bug
>          Components: segment-tar
>            Reporter: Michael Dürig
>            Assignee: Francesco Mari
>              Labels: tools
>             Fix For: Segment Tar 0.0.16
>         Attachments: OAK-2498-01.patch
> According to the [documentation | 
> http://jackrabbit.apache.org/oak/docs/nodestore/segmentmk.html] the root 
> record references in a segment header provide enough context for parsing all 
> records within this segment without any external information. 
> Turns out this is not true: if a root record reference turns e.g. to a list 
> record. The items in that list are record ids of unknown type. So even though 
> those records might live in the same segment, we can't parse them as we don't 
> know their type. 

This message was sent by Atlassian JIRA

Reply via email to