Plan for Hybrid Logical Clocks

[Amit Patel]

Hi everyone,



I'm Amit and I've been picking up on the past work on Hybrid Logical Clocks
(HBASE-14070 ) that was
done by Sai Teja Ranuva last summer. The most recent status of HLC on HBase
can be found here

and a prior document by Enis Soztutar with discussion can be found here
.
I think the effort is at a point where it makes sense to create a public
branch.



Currently the plan is to create a public branch and incrementally add on
functionality and tests. Initially, core HLC would be introduced as a
commit. From there, HLC would then be enabled for just the meta table
(updating the clock on events like region open/close, recovery,
replication, etc). Once we are fully confident that HLC at least works on
the meta table then I anticipate it would be appropriate to merge with the
master branch.



Follow up commits would extend the effort by:



   -

   Addition of a protobuf message called NodeTime that holds timestamp
   corresponding to a send event between nodes.
   -

   Addition of NodeTime as a field to messages sent between nodes for
   messages like requests and responses for region open/close.
   -

   Addition of integration tests that verify HLC works correctly with
   recovery, replication, region open, and region close.
   -

   Later enabling HLC on user tables (main barrier are tests that rely on
   client-side setting of timestamps for table mutations which would no longer
   be allowed on an HLC table).



Work that I have implemented under the outstanding work

as a part of getting HLC working for user tables includes:



   -

   Mapping of time ranges from client GETs
   -

   Disallowing clients to set timestamps for HLC and System Monotonic tables



Feel free to chime in with any comments, suggestions, or other input.

-- 
Amit

Re: Plan for Hybrid Logical Clocks

[Stack]

On Thu, Jun 22, 2017 at 5:34 PM, Amit Patel  wrote:

> Hi everyone,
>
>
>
> I'm Amit and I've been picking up on the past work on Hybrid Logical Clocks
> (HBASE-14070 ) that was
> done by Sai Teja Ranuva last summer. The most recent status of HLC on HBase
> can be found here
>  NuF8TbQkZiZCwFhbHZ7JXRjg3oA/edit#>
> and a prior document by Enis Soztutar with discussion can be found here
>  bXy8P9h6kWC05Bhw/edit#>.
> I think the effort is at a point where it makes sense to create a public
> branch.
>
>
>
> Currently the plan is to create a public branch and incrementally add on
> functionality and tests. Initially, core HLC would be introduced as a
> commit. From there, HLC would then be enabled for just the meta table
> (updating the clock on events like region open/close, recovery,
> replication, etc). Once we are fully confident that HLC at least works on
> the meta table then I anticipate it would be appropriate to merge with the
> master branch.
>
>
>
> Follow up commits would extend the effort by:
>
>
>
>-
>
>Addition of a protobuf message called NodeTime that holds timestamp
>corresponding to a send event between nodes.
>-
>
>Addition of NodeTime as a field to messages sent between nodes for
>messages like requests and responses for region open/close.
>-
>
>Addition of integration tests that verify HLC works correctly with
>recovery, replication, region open, and region close.
>-
>
>Later enabling HLC on user tables (main barrier are tests that rely on
>client-side setting of timestamps for table mutations which would no
> longer
>be allowed on an HLC table).
>
>
>
> Work that I have implemented under the outstanding work
>  NuF8TbQkZiZCwFhbHZ7JXRjg3oA/edit#heading=h.8x5d8iakmo8b>
> as a part of getting HLC working for user tables includes:
>
>
>
>-
>
>Mapping of time ranges from client GETs
>-
>
>Disallowing clients to set timestamps for HLC and System Monotonic
> tables
>
>
>
> Feel free to chime in with any comments, suggestions, or other input.
>


Sounds great to me (smile).
Go Amit!
S




> --
> Amit
>

Re: Plan for Hybrid Logical Clocks

[Sean Busbey]

It'd be great to see this land. You only mention the master branch as
a merge target; is the assumption that this would only be a HBase 3.0+
addition?

On Thu, Jun 22, 2017 at 7:34 PM, Amit Patel  wrote:
> Hi everyone,
>
>
>
> I'm Amit and I've been picking up on the past work on Hybrid Logical Clocks
> (HBASE-14070 ) that was
> done by Sai Teja Ranuva last summer. The most recent status of HLC on HBase
> can be found here
> 
> and a prior document by Enis Soztutar with discussion can be found here
> .
> I think the effort is at a point where it makes sense to create a public
> branch.
>
>
>
> Currently the plan is to create a public branch and incrementally add on
> functionality and tests. Initially, core HLC would be introduced as a
> commit. From there, HLC would then be enabled for just the meta table
> (updating the clock on events like region open/close, recovery,
> replication, etc). Once we are fully confident that HLC at least works on
> the meta table then I anticipate it would be appropriate to merge with the
> master branch.
>
>
>
> Follow up commits would extend the effort by:
>
>
>
>-
>
>Addition of a protobuf message called NodeTime that holds timestamp
>corresponding to a send event between nodes.
>-
>
>Addition of NodeTime as a field to messages sent between nodes for
>messages like requests and responses for region open/close.
>-
>
>Addition of integration tests that verify HLC works correctly with
>recovery, replication, region open, and region close.
>-
>
>Later enabling HLC on user tables (main barrier are tests that rely on
>client-side setting of timestamps for table mutations which would no longer
>be allowed on an HLC table).
>
>
>
> Work that I have implemented under the outstanding work
> 
> as a part of getting HLC working for user tables includes:
>
>
>
>-
>
>Mapping of time ranges from client GETs
>-
>
>Disallowing clients to set timestamps for HLC and System Monotonic tables
>
>
>
> Feel free to chime in with any comments, suggestions, or other input.
>
> --
> Amit

Re: Plan for Hybrid Logical Clocks

[Amit Patel]

For now, yes, that is the plan. However, if it is done before HBase 2.0.0
beta and if it is proven to be solid, then we can discuss bringing it back
to 2.0.0.

On Fri, Jun 23, 2017 at 6:12 AM, Sean Busbey  wrote:

> It'd be great to see this land. You only mention the master branch as
> a merge target; is the assumption that this would only be a HBase 3.0+
> addition?
>
> On Thu, Jun 22, 2017 at 7:34 PM, Amit Patel 
> wrote:
> > Hi everyone,
> >
> >
> >
> > I'm Amit and I've been picking up on the past work on Hybrid Logical
> Clocks
> > (HBASE-14070 ) that
> was
> > done by Sai Teja Ranuva last summer. The most recent status of HLC on
> HBase
> > can be found here
> >  NuF8TbQkZiZCwFhbHZ7JXRjg3oA/edit#>
> > and a prior document by Enis Soztutar with discussion can be found here
> >  bXy8P9h6kWC05Bhw/edit#>.
> > I think the effort is at a point where it makes sense to create a public
> > branch.
> >
> >
> >
> > Currently the plan is to create a public branch and incrementally add on
> > functionality and tests. Initially, core HLC would be introduced as a
> > commit. From there, HLC would then be enabled for just the meta table
> > (updating the clock on events like region open/close, recovery,
> > replication, etc). Once we are fully confident that HLC at least works on
> > the meta table then I anticipate it would be appropriate to merge with
> the
> > master branch.
> >
> >
> >
> > Follow up commits would extend the effort by:
> >
> >
> >
> >-
> >
> >Addition of a protobuf message called NodeTime that holds timestamp
> >corresponding to a send event between nodes.
> >-
> >
> >Addition of NodeTime as a field to messages sent between nodes for
> >messages like requests and responses for region open/close.
> >-
> >
> >Addition of integration tests that verify HLC works correctly with
> >recovery, replication, region open, and region close.
> >-
> >
> >Later enabling HLC on user tables (main barrier are tests that rely on
> >client-side setting of timestamps for table mutations which would no
> longer
> >be allowed on an HLC table).
> >
> >
> >
> > Work that I have implemented under the outstanding work
> >  NuF8TbQkZiZCwFhbHZ7JXRjg3oA/edit#heading=h.8x5d8iakmo8b>
> > as a part of getting HLC working for user tables includes:
> >
> >
> >
> >-
> >
> >Mapping of time ranges from client GETs
> >-
> >
> >Disallowing clients to set timestamps for HLC and System Monotonic
> tables
> >
> >
> >
> > Feel free to chime in with any comments, suggestions, or other input.
> >
> > --
> > Amit
>



-- 
Amit

Re: Plan for Hybrid Logical Clocks

[Enis Söztutar]

The plan looks good.

>However, if it is done before HBase 2.0.0 beta and if it is proven to be
solid, then we can discuss bringing it back to 2.0.0.
It will be pretty good to have at least the base in 2.0, so that we can
later enable HLC for user tables in later releases. Of course, it depends
sufficient testing.

Enis

On Fri, Jun 23, 2017 at 10:03 AM, Amit Patel 
wrote:

> For now, yes, that is the plan. However, if it is done before HBase 2.0.0
> beta and if it is proven to be solid, then we can discuss bringing it back
> to 2.0.0.
>
> On Fri, Jun 23, 2017 at 6:12 AM, Sean Busbey  wrote:
>
> > It'd be great to see this land. You only mention the master branch as
> > a merge target; is the assumption that this would only be a HBase 3.0+
> > addition?
> >
> > On Thu, Jun 22, 2017 at 7:34 PM, Amit Patel 
> > wrote:
> > > Hi everyone,
> > >
> > >
> > >
> > > I'm Amit and I've been picking up on the past work on Hybrid Logical
> > Clocks
> > > (HBASE-14070 ) that
> > was
> > > done by Sai Teja Ranuva last summer. The most recent status of HLC on
> > HBase
> > > can be found here
> > >  > NuF8TbQkZiZCwFhbHZ7JXRjg3oA/edit#>
> > > and a prior document by Enis Soztutar with discussion can be found here
> > >  > bXy8P9h6kWC05Bhw/edit#>.
> > > I think the effort is at a point where it makes sense to create a
> public
> > > branch.
> > >
> > >
> > >
> > > Currently the plan is to create a public branch and incrementally add
> on
> > > functionality and tests. Initially, core HLC would be introduced as a
> > > commit. From there, HLC would then be enabled for just the meta table
> > > (updating the clock on events like region open/close, recovery,
> > > replication, etc). Once we are fully confident that HLC at least works
> on
> > > the meta table then I anticipate it would be appropriate to merge with
> > the
> > > master branch.
> > >
> > >
> > >
> > > Follow up commits would extend the effort by:
> > >
> > >
> > >
> > >-
> > >
> > >Addition of a protobuf message called NodeTime that holds timestamp
> > >corresponding to a send event between nodes.
> > >-
> > >
> > >Addition of NodeTime as a field to messages sent between nodes for
> > >messages like requests and responses for region open/close.
> > >-
> > >
> > >Addition of integration tests that verify HLC works correctly with
> > >recovery, replication, region open, and region close.
> > >-
> > >
> > >Later enabling HLC on user tables (main barrier are tests that rely
> on
> > >client-side setting of timestamps for table mutations which would no
> > longer
> > >be allowed on an HLC table).
> > >
> > >
> > >
> > > Work that I have implemented under the outstanding work
> > >  > NuF8TbQkZiZCwFhbHZ7JXRjg3oA/edit#heading=h.8x5d8iakmo8b>
> > > as a part of getting HLC working for user tables includes:
> > >
> > >
> > >
> > >-
> > >
> > >Mapping of time ranges from client GETs
> > >-
> > >
> > >Disallowing clients to set timestamps for HLC and System Monotonic
> > tables
> > >
> > >
> > >
> > > Feel free to chime in with any comments, suggestions, or other input.
> > >
> > > --
> > > Amit
> >
>
>
>
> --
> Amit
>

Re: Plan for Hybrid Logical Clocks

[Amit Patel]

Hi everyone,

As an update, the core HLC work has been pushed out as an initial commit to
a public branch called HBASE-14070.HLC
.
Details regarding the core HLC commit can be viewed at HBASE-18305
. The next step will be
another commit on that branch that would undo using the master's timestamp
for meta updates. The progress for this task will be tracked in HBASE-18328
.

On Fri, Jun 23, 2017 at 10:38 AM, Enis Söztutar  wrote:

> The plan looks good.
>
> >However, if it is done before HBase 2.0.0 beta and if it is proven to be
> solid, then we can discuss bringing it back to 2.0.0.
> It will be pretty good to have at least the base in 2.0, so that we can
> later enable HLC for user tables in later releases. Of course, it depends
> sufficient testing.
>
> Enis
>
> On Fri, Jun 23, 2017 at 10:03 AM, Amit Patel 
> wrote:
>
> > For now, yes, that is the plan. However, if it is done before HBase 2.0.0
> > beta and if it is proven to be solid, then we can discuss bringing it
> back
> > to 2.0.0.
> >
> > On Fri, Jun 23, 2017 at 6:12 AM, Sean Busbey  wrote:
> >
> > > It'd be great to see this land. You only mention the master branch as
> > > a merge target; is the assumption that this would only be a HBase 3.0+
> > > addition?
> > >
> > > On Thu, Jun 22, 2017 at 7:34 PM, Amit Patel 
> > > wrote:
> > > > Hi everyone,
> > > >
> > > >
> > > >
> > > > I'm Amit and I've been picking up on the past work on Hybrid Logical
> > > Clocks
> > > > (HBASE-14070 )
> that
> > > was
> > > > done by Sai Teja Ranuva last summer. The most recent status of HLC on
> > > HBase
> > > > can be found here
> > > >  > > NuF8TbQkZiZCwFhbHZ7JXRjg3oA/edit#>
> > > > and a prior document by Enis Soztutar with discussion can be found
> here
> > > >  > > bXy8P9h6kWC05Bhw/edit#>.
> > > > I think the effort is at a point where it makes sense to create a
> > public
> > > > branch.
> > > >
> > > >
> > > >
> > > > Currently the plan is to create a public branch and incrementally add
> > on
> > > > functionality and tests. Initially, core HLC would be introduced as a
> > > > commit. From there, HLC would then be enabled for just the meta table
> > > > (updating the clock on events like region open/close, recovery,
> > > > replication, etc). Once we are fully confident that HLC at least
> works
> > on
> > > > the meta table then I anticipate it would be appropriate to merge
> with
> > > the
> > > > master branch.
> > > >
> > > >
> > > >
> > > > Follow up commits would extend the effort by:
> > > >
> > > >
> > > >
> > > >-
> > > >
> > > >Addition of a protobuf message called NodeTime that holds
> timestamp
> > > >corresponding to a send event between nodes.
> > > >-
> > > >
> > > >Addition of NodeTime as a field to messages sent between nodes for
> > > >messages like requests and responses for region open/close.
> > > >-
> > > >
> > > >Addition of integration tests that verify HLC works correctly with
> > > >recovery, replication, region open, and region close.
> > > >-
> > > >
> > > >Later enabling HLC on user tables (main barrier are tests that
> rely
> > on
> > > >client-side setting of timestamps for table mutations which would
> no
> > > longer
> > > >be allowed on an HLC table).
> > > >
> > > >
> > > >
> > > > Work that I have implemented under the outstanding work
> > > >  > > NuF8TbQkZiZCwFhbHZ7JXRjg3oA/edit#heading=h.8x5d8iakmo8b>
> > > > as a part of getting HLC working for user tables includes:
> > > >
> > > >
> > > >
> > > >-
> > > >
> > > >Mapping of time ranges from client GETs
> > > >-
> > > >
> > > >Disallowing clients to set timestamps for HLC and System Monotonic
> > > tables
> > > >
> > > >
> > > >
> > > > Feel free to chime in with any comments, suggestions, or other input.
> > > >
> > > > --
> > > > Amit
> > >
> >
> >
> >
> > --
> > Amit
> >
>



-- 
Amit