Cache sharing between siblings

[Veiko Kukk]

Hi,

I'm new to ATS, so far we've been using Squid, but due to it's instability
are moving away.

I'm interested in avoiding regional cache duplication in CDN system.
Currently, we have several Squid servers working as siblings and only
pulling from origin if none of them has the object in cache, if sibling has
object in cache, it's pulled from sibling but not stored locally. Only new
content fetched from origin is store locally. Sibling lookup is done via
ICP and has been working quite well so far.

Is similar setup possible with ATS?
I found no information in documentation how to configure siblings (only
parent configuration is documented). Could you point me to right direction?

Regards,
Veiko

[Veiko Kukk]

Hi again,

Are there any better lists/forums to ask such questions?

Regards,
Veiko

Post by Veiko Kukk
Hi,
I'm new to ATS, so far we've been using Squid, but due to it's instability
are moving away.
I'm interested in avoiding regional cache duplication in CDN system.
Currently, we have several Squid servers working as siblings and only
pulling from origin if none of them has the object in cache, if sibling has
object in cache, it's pulled from sibling but not stored locally. Only new
content fetched from origin is store locally. Sibling lookup is done via
ICP and has been working quite well so far.
Is similar setup possible with ATS?
I found no information in documentation how to configure siblings (only
parent configuration is documented). Could you point me to right direction?
Regards,
Veiko

[David Carlin]

ATS doesn't have this functionality currently. I know a couple companies
are working on it, but nothing to contribute to open source yet as far as I
know.

David

Post by Veiko Kukk
Hi again,
Are there any better lists/forums to ask such questions?
Regards,
Veiko

Post by Veiko Kukk
Hi,
I'm new to ATS, so far we've been using Squid, but due to it's
instability are moving away.
I'm interested in avoiding regional cache duplication in CDN system.
Currently, we have several Squid servers working as siblings and only
pulling from origin if none of them has the object in cache, if sibling has
object in cache, it's pulled from sibling but not stored locally. Only new
content fetched from origin is store locally. Sibling lookup is done via
ICP and has been working quite well so far.
Is similar setup possible with ATS?
I found no information in documentation how to configure siblings (only
parent configuration is documented). Could you point me to right direction?
Regards,
Veiko

[Alan Carroll]

Sorry for the delay, and thanks Dave for answering. This is a feature under
active development. Oath internally uses a plugin that implements CARP for
sharing across a pod. We are trying to open source it but it's currently
too flaky to support externally. We are working on improving it to the
point where it can be open sourced without embarrassment. In the longer
run, however, there is a "Layer 7 Routing" project (also known as "Next
Hop") which you can track here -
https://github.com/apache/trafficserver/projects/7

Post by David Carlin
ATS doesn't have this functionality currently. I know a couple companies
are working on it, but nothing to contribute to open source yet as far as I
know.
David

Post by Veiko Kukk
Hi again,
Are there any better lists/forums to ask such questions?
Regards,
Veiko

Post by Veiko Kukk
Hi,
I'm new to ATS, so far we've been using Squid, but due to it's
instability are moving away.
I'm interested in avoiding regional cache duplication in CDN system.
Currently, we have several Squid servers working as siblings and only
pulling from origin if none of them has the object in cache, if sibling has
object in cache, it's pulled from sibling but not stored locally. Only new
content fetched from origin is store locally. Sibling lookup is done via
ICP and has been working quite well so far.
Is similar setup possible with ATS?
I found no information in documentation how to configure siblings (only
parent configuration is documented). Could you point me to right direction?
Regards,
Veiko

[Jeremy Payne]

I've been thinking about this, but you can use current parent
functionality to perform local node(edge) object lookups.
It's a hack and requires careful configuration of your remap and
parent configuration so to not end up in a cache loop
or end up storing duplicate copies.. but it is a doable workaround in
the absence of some type of clustering(CARP, etc).

Post by Alan Carroll
Sorry for the delay, and thanks Dave for answering. This is a feature under
active development. Oath internally uses a plugin that implements CARP for
sharing across a pod. We are trying to open source it but it's currently too
flaky to support externally. We are working on improving it to the point
where it can be open sourced without embarrassment. In the longer run,
however, there is a "Layer 7 Routing" project (also known as "Next Hop")
which you can track here -
https://github.com/apache/trafficserver/projects/7

Post by David Carlin
ATS doesn't have this functionality currently. I know a couple companies
are working on it, but nothing to contribute to open source yet as far as I
know.
David

Post by Veiko Kukk
Hi again,
Are there any better lists/forums to ask such questions?
Regards,
Veiko

Post by Veiko Kukk
Hi,
I'm new to ATS, so far we've been using Squid, but due to it's
instability are moving away.
I'm interested in avoiding regional cache duplication in CDN system.
Currently, we have several Squid servers working as siblings and only
pulling from origin if none of them has the object in cache, if sibling has
object in cache, it's pulled from sibling but not stored locally. Only new
content fetched from origin is store locally. Sibling lookup is done via ICP
and has been working quite well so far.
Is similar setup possible with ATS?
I found no information in documentation how to configure siblings (only
parent configuration is documented). Could you point me to right direction?
Regards,
Veiko

[Jeremy Payne]

btw... i'll add a lesson learned post to my wiki..
give me a couple days to clean up lab details.. :-)

Post by Jeremy Payne
I've been thinking about this, but you can use current parent
functionality to perform local node(edge) object lookups.
It's a hack and requires careful configuration of your remap and
parent configuration so to not end up in a cache loop
or end up storing duplicate copies.. but it is a doable workaround in
the absence of some type of clustering(CARP, etc).

Post by Alan Carroll
Sorry for the delay, and thanks Dave for answering. This is a feature under
active development. Oath internally uses a plugin that implements CARP for
sharing across a pod. We are trying to open source it but it's currently too
flaky to support externally. We are working on improving it to the point
where it can be open sourced without embarrassment. In the longer run,
however, there is a "Layer 7 Routing" project (also known as "Next Hop")
which you can track here -
https://github.com/apache/trafficserver/projects/7

Post by David Carlin
ATS doesn't have this functionality currently. I know a couple companies
are working on it, but nothing to contribute to open source yet as far as I
know.
David

Post by Veiko Kukk
Hi again,
Are there any better lists/forums to ask such questions?
Regards,
Veiko

Post by Veiko Kukk
Hi,
I'm new to ATS, so far we've been using Squid, but due to it's
instability are moving away.
I'm interested in avoiding regional cache duplication in CDN system.
Currently, we have several Squid servers working as siblings and only
pulling from origin if none of them has the object in cache, if sibling has
object in cache, it's pulled from sibling but not stored locally. Only new
content fetched from origin is store locally. Sibling lookup is done via ICP
and has been working quite well so far.
Is similar setup possible with ATS?
I found no information in documentation how to configure siblings (only
parent configuration is documented). Could you point me to right direction?
Regards,
Veiko