C++ accumulo client --> native clients for Python, Go, Ruby etc
Accumulo Developers, We're trying to boost throughput of non-Java tools with Accumulo. It seems that the lowest hanging fruit is to stop using the thrift proxy. Per discussion about Python and thrift proxy in the users list [1], I'm wondering if anyone is interested in helping with a native C++ client? There is a start on one here [2]. We could offer a bounty or maybe make a consulting project depending who is interested in it. We also looked at trying to run a separate thrift proxy for every worker thread or process. With many cores on a box, eg 32, it just doesn't seem practical to run that many proxies, even if they all run on a single JVM. We'd be glad to hear ideas on that front too. A potentially big benefit of making a proper C++ accumulo client is that it is straightforward to expose native interfaces in Python (via pyObject), Go [3], Ruby [4], and other languages. Thanks for any advice, pointers, interest. John 1-- http://www.mail-archive.com/user@accumulo.apache.org/msg03999.html 2-- https://github.com/phrocker/apeirogon 3-- http://golang.org/cmd/cgo/ 4-- https://www.amberbit.com/blog/2014/6/12/calling-c-cpp-from-ruby/ Sent from +1-617-899-2066
Re: C++ accumulo client --> native clients for Python, Go, Ruby etc
It'd be really cool to see a C++ client -- fully implemented or not. The increased performance via other languages like you said would be really nice, but I'd also be curious to see how the server characteristics change when the client might be sending data at a much faster rate. My C++ is super rusty these days, but I'd be happy to help out any devs who can spearhead the effort :) John R. Frank wrote: Accumulo Developers, We're trying to boost throughput of non-Java tools with Accumulo. It seems that the lowest hanging fruit is to stop using the thrift proxy. Per discussion about Python and thrift proxy in the users list [1], I'm wondering if anyone is interested in helping with a native C++ client? There is a start on one here [2]. We could offer a bounty or maybe make a consulting project depending who is interested in it. We also looked at trying to run a separate thrift proxy for every worker thread or process. With many cores on a box, eg 32, it just doesn't seem practical to run that many proxies, even if they all run on a single JVM. We'd be glad to hear ideas on that front too. A potentially big benefit of making a proper C++ accumulo client is that it is straightforward to expose native interfaces in Python (via pyObject), Go [3], Ruby [4], and other languages. Thanks for any advice, pointers, interest. John 1-- http://www.mail-archive.com/user@accumulo.apache.org/msg03999.html 2-- https://github.com/phrocker/apeirogon 3-- http://golang.org/cmd/cgo/ 4-- https://www.amberbit.com/blog/2014/6/12/calling-c-cpp-from-ruby/ Sent from +1-617-899-2066
Re: C++ accumulo client --> native clients for Python, Go, Ruby etc
I'm all for this- though I'm curious to know the thoughts about maintenance and the design. Are we going to use thrift to tie the C++ client calls into the server-side components? Is that going to be maintained through a separate effort or is the plan to have the Accumulo community officially support it? On Mon, Oct 6, 2014 at 2:34 PM, Josh Elser wrote: > It'd be really cool to see a C++ client -- fully implemented or not. The > increased performance via other languages like you said would be really > nice, but I'd also be curious to see how the server characteristics change > when the client might be sending data at a much faster rate. > > My C++ is super rusty these days, but I'd be happy to help out any devs > who can spearhead the effort :) > > > John R. Frank wrote: > >> Accumulo Developers, >> >> We're trying to boost throughput of non-Java tools with Accumulo. It >> seems that the lowest hanging fruit is to stop using the thrift proxy. Per >> discussion about Python and thrift proxy in the users list [1], I'm >> wondering if anyone is interested in helping with a native C++ client? >> There is a start on one here [2]. We could offer a bounty or maybe make a >> consulting project depending who is interested in it. >> >> We also looked at trying to run a separate thrift proxy for every worker >> thread or process. With many cores on a box, eg 32, it just doesn't seem >> practical to run that many proxies, even if they all run on a single JVM. >> We'd be glad to hear ideas on that front too. >> >> A potentially big benefit of making a proper C++ accumulo client is that >> it is straightforward to expose native interfaces in Python (via pyObject), >> Go [3], Ruby [4], and other languages. >> >> Thanks for any advice, pointers, interest. >> >> John >> >> >> 1-- http://www.mail-archive.com/user@accumulo.apache.org/msg03999.html >> >> 2-- >> https://github.com/phrocker/apeirogon >> >> 3-- http://golang.org/cmd/cgo/ >> >> 4-- https://www.amberbit.com/blog/2014/6/12/calling-c-cpp-from-ruby/ >> >> >> Sent from +1-617-899-2066 >> >
Re: C++ accumulo client --> native clients for Python, Go, Ruby etc
There is an initial start on a C++ client here: https://github.com/phrocker/apeirogon/issues/1#issuecomment-57958812 Looks like it is starting to use the underlying thrift interfaces, just like the java client. It would be great to get this supported in core accumulo. My company would be interested in helping to support work on this. To get it jump started, we could offer a bounty or maybe focused consulting engagement, and then longer term we could help maintain the code (since we'd be using it.) jrf On Mon, 6 Oct 2014, Corey Nolet wrote: I'm all for this- though I'm curious to know the thoughts about maintenance and the design. Are we going to use thrift to tie the C++ client calls into the server-side components? Is that going to be maintained through a separate effort or is the plan to have the Accumulo community officially support it? On Mon, Oct 6, 2014 at 2:34 PM, Josh Elser wrote: It'd be really cool to see a C++ client -- fully implemented or not. The increased performance via other languages like you said would be really nice, but I'd also be curious to see how the server characteristics change when the client might be sending data at a much faster rate. My C++ is super rusty these days, but I'd be happy to help out any devs who can spearhead the effort :) John R. Frank wrote: Accumulo Developers, We're trying to boost throughput of non-Java tools with Accumulo. It seems that the lowest hanging fruit is to stop using the thrift proxy. Per discussion about Python and thrift proxy in the users list [1], I'm wondering if anyone is interested in helping with a native C++ client? There is a start on one here [2]. We could offer a bounty or maybe make a consulting project depending who is interested in it. We also looked at trying to run a separate thrift proxy for every worker thread or process. With many cores on a box, eg 32, it just doesn't seem practical to run that many proxies, even if they all run on a single JVM. We'd be glad to hear ideas on that front too. A potentially big benefit of making a proper C++ accumulo client is that it is straightforward to expose native interfaces in Python (via pyObject), Go [3], Ruby [4], and other languages. Thanks for any advice, pointers, interest. John 1-- http://www.mail-archive.com/user@accumulo.apache.org/msg03999.html 2-- https://github.com/phrocker/apeirogon 3-- http://golang.org/cmd/cgo/ 4-- https://www.amberbit.com/blog/2014/6/12/calling-c-cpp-from-ruby/ Sent from +1-617-899-2066
Re: C++ accumulo client --> native clients for Python, Go, Ruby etc
How far away from the theoretical maximum rate is the thrift protocol? What kind of gain is expected from the native C++ approach? On Sat, Oct 4, 2014 at 12:56 PM, John R. Frank wrote: > Accumulo Developers, > > We're trying to boost throughput of non-Java tools with Accumulo. It seems > that the lowest hanging fruit is to stop using the thrift proxy. Per > discussion about Python and thrift proxy in the users list [1], I'm wondering > if anyone is interested in helping with a native C++ client? There is a > start on one here [2]. We could offer a bounty or maybe make a consulting > project depending who is interested in it. > > We also looked at trying to run a separate thrift proxy for every worker > thread or process. With many cores on a box, eg 32, it just doesn't seem > practical to run that many proxies, even if they all run on a single JVM. > We'd be glad to hear ideas on that front too. > > A potentially big benefit of making a proper C++ accumulo client is that it > is straightforward to expose native interfaces in Python (via pyObject), Go > [3], Ruby [4], and other languages. > > Thanks for any advice, pointers, interest. > > John > > > 1-- http://www.mail-archive.com/user@accumulo.apache.org/msg03999.html > > 2-- > https://github.com/phrocker/apeirogon > > 3-- http://golang.org/cmd/cgo/ > > 4-- https://www.amberbit.com/blog/2014/6/12/calling-c-cpp-from-ruby/ > > > Sent from +1-617-899-2066
Re: C++ accumulo client --> native clients for Python, Go, Ruby etc
Yeah the reason ask about having the community support it is because the thrift interfaces are all internal workings, thus, code other than the client API should probably be maintained internally. This will add a slight overhead to development and maintenance of these thrift services because more implementations will now need to be kept up to date and modified when changes need to be implemented. Not saying it should be a showstopper, just pointing it out. On Mon, Oct 6, 2014 at 4:15 PM, David Medinets wrote: > How far away from the theoretical maximum rate is the thrift protocol? > What kind of gain is expected from the native C++ approach? > > On Sat, Oct 4, 2014 at 12:56 PM, John R. Frank wrote: > > Accumulo Developers, > > > > We're trying to boost throughput of non-Java tools with Accumulo. It > seems that the lowest hanging fruit is to stop using the thrift proxy. Per > discussion about Python and thrift proxy in the users list [1], I'm > wondering if anyone is interested in helping with a native C++ client? > There is a start on one here [2]. We could offer a bounty or maybe make a > consulting project depending who is interested in it. > > > > We also looked at trying to run a separate thrift proxy for every worker > thread or process. With many cores on a box, eg 32, it just doesn't seem > practical to run that many proxies, even if they all run on a single JVM. > We'd be glad to hear ideas on that front too. > > > > A potentially big benefit of making a proper C++ accumulo client is that > it is straightforward to expose native interfaces in Python (via pyObject), > Go [3], Ruby [4], and other languages. > > > > Thanks for any advice, pointers, interest. > > > > John > > > > > > 1-- http://www.mail-archive.com/user@accumulo.apache.org/msg03999.html > > > > 2-- > > https://github.com/phrocker/apeirogon > > > > 3-- http://golang.org/cmd/cgo/ > > > > 4-- https://www.amberbit.com/blog/2014/6/12/calling-c-cpp-from-ruby/ > > > > > > Sent from +1-617-899-2066 >
Re: C++ accumulo client --> native clients for Python, Go, Ruby etc
Two kinds of gains: 1) single client throughput: the extra RPC hop through the proxy deserializes and then reserializes the messages. With the proxy running locally the extra network hop is less of an issue. This was discussed on the user list (see link earlier in this thread), and 5x slow down was suggested as a possible swag estimate. 2) cluster management complexity: it's clearly best to have the proxy local to the workers, but if you have a worker on every core of a large box (eg 32), then having a single proxy on each worker machine becomes a bottleneck. Running many proxies on a single JVM is the next thing we could try to improve this --- having a native client seems preferable. Comments? jrf > On Oct 6, 2014, at 4:15 PM, David Medinets wrote: > > How far away from the theoretical maximum rate is the thrift protocol? > What kind of gain is expected from the native C++ approach? > >> On Sat, Oct 4, 2014 at 12:56 PM, John R. Frank wrote: >> Accumulo Developers, >> >> We're trying to boost throughput of non-Java tools with Accumulo. It seems >> that the lowest hanging fruit is to stop using the thrift proxy. Per >> discussion about Python and thrift proxy in the users list [1], I'm >> wondering if anyone is interested in helping with a native C++ client? >> There is a start on one here [2]. We could offer a bounty or maybe make a >> consulting project depending who is interested in it. >> >> We also looked at trying to run a separate thrift proxy for every worker >> thread or process. With many cores on a box, eg 32, it just doesn't seem >> practical to run that many proxies, even if they all run on a single JVM. >> We'd be glad to hear ideas on that front too. >> >> A potentially big benefit of making a proper C++ accumulo client is that it >> is straightforward to expose native interfaces in Python (via pyObject), Go >> [3], Ruby [4], and other languages. >> >> Thanks for any advice, pointers, interest. >> >> John >> >> >> 1-- http://www.mail-archive.com/user@accumulo.apache.org/msg03999.html >> >> 2-- >> https://github.com/phrocker/apeirogon >> >> 3-- http://golang.org/cmd/cgo/ >> >> 4-- https://www.amberbit.com/blog/2014/6/12/calling-c-cpp-from-ruby/ >> >> >> Sent from +1-617-899-2066
Re: C++ accumulo client --> native clients for Python, Go, Ruby etc
John, Do you have any performance numbers that you can share around your use of the existing proxy solution? One of the reasons that Thrift is performant for Accumulo is that messages are batched by the client library and sent over a smaller number of RPCs. A C++ client will also need to have mechanisms like the BatchWriter to get the best performance. Also, it may be possible to make the proxy faster by batching more data into the update call. Adam On Mon, Oct 6, 2014 at 5:20 PM, John R. Frank wrote: > Two kinds of gains: > > 1) single client throughput: the extra RPC hop through the proxy > deserializes and then reserializes the messages. With the proxy running > locally the extra network hop is less of an issue. This was discussed on the > user list (see link earlier in this thread), and 5x slow down was suggested > as a possible swag estimate. > > 2) cluster management complexity: it's clearly best to have the proxy local > to the workers, but if you have a worker on every core of a large box (eg > 32), then having a single proxy on each worker machine becomes a bottleneck. > Running many proxies on a single JVM is the next thing we could try to > improve this --- having a native client seems preferable. > > > Comments? > > jrf > > >> On Oct 6, 2014, at 4:15 PM, David Medinets wrote: >> >> How far away from the theoretical maximum rate is the thrift protocol? >> What kind of gain is expected from the native C++ approach? >> >>> On Sat, Oct 4, 2014 at 12:56 PM, John R. Frank wrote: >>> Accumulo Developers, >>> >>> We're trying to boost throughput of non-Java tools with Accumulo. It seems >>> that the lowest hanging fruit is to stop using the thrift proxy. Per >>> discussion about Python and thrift proxy in the users list [1], I'm >>> wondering if anyone is interested in helping with a native C++ client? >>> There is a start on one here [2]. We could offer a bounty or maybe make a >>> consulting project depending who is interested in it. >>> >>> We also looked at trying to run a separate thrift proxy for every worker >>> thread or process. With many cores on a box, eg 32, it just doesn't seem >>> practical to run that many proxies, even if they all run on a single JVM. >>> We'd be glad to hear ideas on that front too. >>> >>> A potentially big benefit of making a proper C++ accumulo client is that it >>> is straightforward to expose native interfaces in Python (via pyObject), Go >>> [3], Ruby [4], and other languages. >>> >>> Thanks for any advice, pointers, interest. >>> >>> John >>> >>> >>> 1-- http://www.mail-archive.com/user@accumulo.apache.org/msg03999.html >>> >>> 2-- >>> https://github.com/phrocker/apeirogon >>> >>> 3-- http://golang.org/cmd/cgo/ >>> >>> 4-- https://www.amberbit.com/blog/2014/6/12/calling-c-cpp-from-ruby/ >>> >>> >>> Sent from +1-617-899-2066
Re: C++ accumulo client --> native clients for Python, Go, Ruby etc
We're running a experiment with multiple local proxies to try to better understand the batching and bottlenecking issues. Will report back. jrf