[
https://issues.apache.org/jira/browse/HBASE-8927?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Lars Hofhansl updated HBASE-8927:
---------------------------------
Attachment: 8927-WIP-TEST.txt
Here's a _work in progress_ just need to park it here.
Observations:
* timestamp_multiplier is weird. Who is going to understand what that means?
* Here I have used timestamp_resolution, but that's weird too. Nobody would
want to lower than 1000 anyway. And unless we track a nominator and denominator
the resolution needs to be dividable by 1000.
* Assumptions about ms are all over the place. Need to scale it everywhere, but
not where we actually need wall time (for example in the WAL keys). Getting
this all right will be tricky.
* In introduced a getNow method to HRegion that can return an appropriate
representation of _now_ for the region.
So I think I want to call it timestamp_resolution, but store it as multiplier
only (i.e. 1 for ms, 1000 for us, 1000000 for ns). Then I also need to
introduce the constant I mentioned.
> Use nano time instead of mili time everywhere
> ---------------------------------------------
>
> Key: HBASE-8927
> URL: https://issues.apache.org/jira/browse/HBASE-8927
> Project: HBase
> Issue Type: Bug
> Reporter: stack
> Labels: phoenix
> Attachments: 8927-WIP-TEST.txt, 8927.txt
>
>
> Less collisions and we are paying the price of a long anyways so might as
> well fill it.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
