[jira] [Commented] (OAK-4095) Include timestamp in journal log entries

["Michael Dürig (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/OAK-4095?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15185549#comment-15185549 ] 

Michael Dürig commented on OAK-4095:
------------------------------------

My preferred solution would be OAK-4103. Until then I'm fine with adding timestamps if we can do it in a reasonably safe way wrt. backwards compatibility. 

> Include timestamp in journal log entries
> ----------------------------------------
>
>                 Key: OAK-4095
>                 URL: https://issues.apache.org/jira/browse/OAK-4095
>             Project: Jackrabbit Oak
>          Issue Type: Improvement
>          Components: segmentmk
>            Reporter: Chetan Mehrotra
>            Priority: Minor
>              Labels: candidate_oak_1_0, candidate_oak_1_2, candidate_oak_1_4
>             Fix For: 1.6
>
>
> Currently the journal log has entries like below. At times while debugging crash or some issue we need to determine the probable root state at some point in the past. 
> {noformat}
> 3dea11bb-bd43-4319-a37d-59df778a7271:260988 root
> a7a509ac-a9d4-4e2c-a0d8-df71ebe123a0:259736 root
> 1d889da9-b41c-4889-a0cd-a9aa9dcc1737:259992 root
> b78e4aa6-ec68-4e70-a364-f04ccbf4c3b3:259964 root
> {noformat}
> Currently there is no way to determine from above log what is the root state wrt time. So we need to workaround that by reading each root state and look for some path which has some time related property. To simplify such case it would be helpful to also include timestamp while adding a journal entry
> {noformat}
> 1d889da9-b41c-4889-a0cd-a9aa9dcc1737:259992 root 1457408708772
> b78e4aa6-ec68-4e70-a364-f04ccbf4c3b3:259964 root 1457408708899
> {noformat}
> *Key points*
> # Timestamp comes at end
> # Such a feature can be enabled without affecting backward compatibility - Just that new entries would have timestamp included
> # {{JournalReader}} - Just reads the first column so would work as is



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)