CF-Persistence Proposal


Michael Maximilien
 

Hi, all,

For the past month, working with colleagues from IBM Research, as well as colleagues from Pivotal, and EMC, we have created the first version of a proposal to add persistent storage to CF applications.

https://docs.google.com/document/d/1A1PVnwB7wdzrWq2ZTjNrDFULlmyTUSsOuWeih8kdUtw/edit?usp=docslist_api

The gist of the proposal is to introduce a special kind of broker that would collaborate with Diego and an external Storage Provider that implements a common interface (SPI) to give CF apps that use this broker, access to persistent storage that not only persist applications crashes but could potentially scale and perform to the ability of the SPI. The goal would be that different SPIs could exist for different types of storage systems such as clustered file systems and others.

We are inviting all interested parties to comment on this proposal and / or this thread. We are hoping for an open discussion and feedback cycle that result in an updated proposal that includes agreed upon steps (architecture, design, goals) for incubating this into CF.

At the end of the day, our aim is to innovate. Expand the platform to more users and make more customers happy.

Hope you join us in this effort. All the best,

Max on behalf of the CF-Persitence working team

dr.max
ibm cloud labs
sillicon valley, ca

Sent from my iPhone


Mike Youngstrom <youngm@...>
 

This is awesome dr. max! Easily the most requested feature in my
enterprise. I'm super excited to see this proposal. I'll give it a review
next week.

Mike

On Fri, Nov 20, 2015 at 4:51 PM, Michael Maximilien <maxim(a)us.ibm.com>
wrote:

Hi, all,

For the past month, working with colleagues from IBM Research, as well as
colleagues from Pivotal, and EMC, we have created the first version of a
proposal to add persistent storage to CF applications.


https://docs.google.com/document/d/1A1PVnwB7wdzrWq2ZTjNrDFULlmyTUSsOuWeih8kdUtw/edit?usp=docslist_api

The gist of the proposal is to introduce a special kind of broker that
would collaborate with Diego and an external Storage Provider that
implements a common interface (SPI) to give CF apps that use this broker,
access to persistent storage that not only persist applications crashes but
could potentially scale and perform to the ability of the SPI. The goal
would be that different SPIs could exist for different types of storage
systems such as clustered file systems and others.

We are inviting all interested parties to comment on this proposal and /
or this thread. We are hoping for an open discussion and feedback cycle
that result in an updated proposal that includes agreed upon steps
(architecture, design, goals) for incubating this into CF.

At the end of the day, our aim is to innovate. Expand the platform to more
users and make more customers happy.

Hope you join us in this effort. All the best,

Max on behalf of the CF-Persitence working team

dr.max
ibm cloud labs
sillicon valley, ca

Sent from my iPhone


James Bayer
 

for reference, this is how kubernetes is categorizing various volumes types
and capabilities [1] as discussed on this tweet [2].

[1]
https://docs.google.com/spreadsheets/d/14-DEUm1HEHLn9yOEBHq9xxR6iZWdO4fYuluNClgEDYw/edit#gid=0&vpid=A1
[2] https://twitter.com/mikemetral/status/672163753291460608

On Fri, Nov 20, 2015 at 9:57 PM Mike Youngstrom <youngm(a)gmail.com> wrote:

This is awesome dr. max! Easily the most requested feature in my
enterprise. I'm super excited to see this proposal. I'll give it a review
next week.

Mike

On Fri, Nov 20, 2015 at 4:51 PM, Michael Maximilien <maxim(a)us.ibm.com>
wrote:

Hi, all,

For the past month, working with colleagues from IBM Research, as well as
colleagues from Pivotal, and EMC, we have created the first version of a
proposal to add persistent storage to CF applications.


https://docs.google.com/document/d/1A1PVnwB7wdzrWq2ZTjNrDFULlmyTUSsOuWeih8kdUtw/edit?usp=docslist_api

The gist of the proposal is to introduce a special kind of broker that
would collaborate with Diego and an external Storage Provider that
implements a common interface (SPI) to give CF apps that use this broker,
access to persistent storage that not only persist applications crashes but
could potentially scale and perform to the ability of the SPI. The goal
would be that different SPIs could exist for different types of storage
systems such as clustered file systems and others.

We are inviting all interested parties to comment on this proposal and /
or this thread. We are hoping for an open discussion and feedback cycle
that result in an updated proposal that includes agreed upon steps
(architecture, design, goals) for incubating this into CF.

At the end of the day, our aim is to innovate. Expand the platform to
more users and make more customers happy.

Hope you join us in this effort. All the best,

Max on behalf of the CF-Persitence working team

dr.max
ibm cloud labs
sillicon valley, ca

Sent from my iPhone


Renat Khasanshyn
 

Oh yeah baby. +1

Andrei, Kirill, please join in. Perhaps we can help with some
contributions. From the first look it probably makes sense to get Ksenia to
review this.

Renat

On Wed, Dec 2, 2015 at 3:59 PM, James Bayer <jbayer(a)pivotal.io> wrote:


for reference, this is how kubernetes is categorizing various volumes
types and capabilities [1] as discussed on this tweet [2].

[1]
https://docs.google.com/spreadsheets/d/14-DEUm1HEHLn9yOEBHq9xxR6iZWdO4fYuluNClgEDYw/edit#gid=0&vpid=A1
[2] https://twitter.com/mikemetral/status/672163753291460608


On Fri, Nov 20, 2015 at 9:57 PM Mike Youngstrom <youngm(a)gmail.com> wrote:

This is awesome dr. max! Easily the most requested feature in my
enterprise. I'm super excited to see this proposal. I'll give it a review
next week.

Mike

On Fri, Nov 20, 2015 at 4:51 PM, Michael Maximilien <maxim(a)us.ibm.com>
wrote:

Hi, all,

For the past month, working with colleagues from IBM Research, as well
as colleagues from Pivotal, and EMC, we have created the first version of a
proposal to add persistent storage to CF applications.


https://docs.google.com/document/d/1A1PVnwB7wdzrWq2ZTjNrDFULlmyTUSsOuWeih8kdUtw/edit?usp=docslist_api

The gist of the proposal is to introduce a special kind of broker that
would collaborate with Diego and an external Storage Provider that
implements a common interface (SPI) to give CF apps that use this broker,
access to persistent storage that not only persist applications crashes but
could potentially scale and perform to the ability of the SPI. The goal
would be that different SPIs could exist for different types of storage
systems such as clustered file systems and others.

We are inviting all interested parties to comment on this proposal and /
or this thread. We are hoping for an open discussion and feedback cycle
that result in an updated proposal that includes agreed upon steps
(architecture, design, goals) for incubating this into CF.

At the end of the day, our aim is to innovate. Expand the platform to
more users and make more customers happy.

Hope you join us in this effort. All the best,

Max on behalf of the CF-Persitence working team

dr.max
ibm cloud labs
sillicon valley, ca

Sent from my iPhone


Michael Maximilien
 

Thanks for sharing James. 




Renat, cool. We look forward to your feedback.




Best,


Max






Sent from Mailbox

On Thu, Dec 3, 2015 at 8:00 AM, James Bayer <jbayer(a)pivotal.io> wrote:

for reference, this is how kubernetes is categorizing various volumes types
and capabilities [1] as discussed on this tweet [2].
[1]
https://docs.google.com/spreadsheets/d/14-DEUm1HEHLn9yOEBHq9xxR6iZWdO4fYuluNClgEDYw/edit#gid=0&vpid=A1
[2] https://twitter.com/mikemetral/status/672163753291460608
On Fri, Nov 20, 2015 at 9:57 PM Mike Youngstrom <youngm(a)gmail.com> wrote:
This is awesome dr. max! Easily the most requested feature in my
enterprise. I'm super excited to see this proposal. I'll give it a review
next week.

Mike

On Fri, Nov 20, 2015 at 4:51 PM, Michael Maximilien <maxim(a)us.ibm.com>
wrote:

Hi, all,

For the past month, working with colleagues from IBM Research, as well as
colleagues from Pivotal, and EMC, we have created the first version of a
proposal to add persistent storage to CF applications.


https://docs.google.com/document/d/1A1PVnwB7wdzrWq2ZTjNrDFULlmyTUSsOuWeih8kdUtw/edit?usp=docslist_api

The gist of the proposal is to introduce a special kind of broker that
would collaborate with Diego and an external Storage Provider that
implements a common interface (SPI) to give CF apps that use this broker,
access to persistent storage that not only persist applications crashes but
could potentially scale and perform to the ability of the SPI. The goal
would be that different SPIs could exist for different types of storage
systems such as clustered file systems and others.

We are inviting all interested parties to comment on this proposal and /
or this thread. We are hoping for an open discussion and feedback cycle
that result in an updated proposal that includes agreed upon steps
(architecture, design, goals) for incubating this into CF.

At the end of the day, our aim is to innovate. Expand the platform to
more users and make more customers happy.

Hope you join us in this effort. All the best,

Max on behalf of the CF-Persitence working team

dr.max
ibm cloud labs
sillicon valley, ca

Sent from my iPhone