[grpc-io] Re: Need basic help with unexplained steps building on Windows [Resurrected]

I'd like to note that this is a CMake issue where cmake picked a wrong version of Windows SDK when building BoringSSL. It looks like your environment hit this hairy issue but I'm afraid that I'm not an expert of cmake either but "CMAKE_SYSTEM_VERSION" option should be specified when you run

[grpc-io] Re: Need basic help with unexplained steps building on Windows [Resurrected]

Your case appears to be relevant to https://gitlab.kitware.com/cmake/cmake/-/issues/16202#note_140259 This is a cmake issue and there is not much gRPC can do. On Tuesday, July 16, 2024 at 11:57:09 AM UTC-7 Kevin Mendel wrote: > CMAKE_VS_WINDOWS_TARGET_PLATFORM_VERSION doesn't seem to work

[grpc-io] Re: grpc 1.51 deb package creation failing with checkinstall with cmake 3.28

As gRPC team doesn't own and maintain gRPC deb package, we have no idea about this issue. You may want to reach out to Debian package maintainer. https://salsa.debian.org/debian/grpc On Thursday, June 6, 2024 at 2:29:51 AM UTC-7 Ramana Reddy wrote: > Hi All, > We are seeing the following error

[grpc-io] Re: Grpc c++ server compiling using was

Building gRPC with emcc is not officially supported at this time. If this is important to you, we welcome your contributions to gRPC. On Thursday, June 13, 2024 at 3:50:22 AM UTC-7 revathi anbumalar wrote: > Getting few errors like below for the sample when compiling using emcc > > 453 |

[grpc-io] Re: Question about dependency on protobuf

Usually gRPC tries to bump the version of Protobuf every time Protobuf has a new release to make two libraries work together. But gRPC can work with older versions of Protobuf in general. You can try using older protobuf with the latest gRPC. Sometimes you might see build errors because of

[grpc-io] Re: GRPC C++ request spent > 2s from client to server

You can use environment variables to get more gRPC logging; See https://github.com/grpc/grpc/blob/master/doc/environment_variables.md I don't think spending 2 seconds to complete a simple RPC look good. On Friday, December 15, 2023 at 12:33:06 PM UTC-8 Frank Yu wrote: > Hi GRPC team, > > I

[grpc-io] Re: c-ares and CVE-2022-4904 in gRPC C# v1.46.x

>From https://github.com/c-ares/c-ares/issues/496, it looks like ares_set_sortlist used to have a security issue but gRPC Core which gRPC C# is using doesn't call this function. Thus, gRPC is not affected by this issue. On Wednesday, November 22, 2023 at 9:56:48 AM UTC-8 Sebas Mendez wrote: >

[grpc-io] Re: gRPC and CVE-2023-44487?

We don't want to share details about how to reproduce it because it would do more harm than good. Action required here to mitigate this is to update gRPC to the version with the fix. On Thursday, October 26, 2023 at 11:48:24 PM UTC-7 yh zhou wrote: > Are there any POCs or steps to reproduce

[grpc-io] gRPC C++ patch releases to fix Cord optimization memory bug

Hi, To fix a memory bug in gRPC C++ that occurs when deserializing absl::Cord data, gRPC C++ has released a few patch versions, 1.55.x through 1.59.x. If you are using gRPC C++ and absl::Cord, please upgrade it to one of these patch versions. -

[grpc-io] Upcoming change on Bazel support

Hi gRPC community, gRPC Core started following the Foundational C++ policy a couple of years ago, which changes the supported platforms and build systems. As a result, gRPC will drop Bazel 5.x support in gRPC 1.60. (Note

[grpc-io] Re: gRPC and CVE-2023-44487?

gRPC C++, Python, and Ruby will soon have a 1.59.2 patch release to address CVE-2023-44487. Thus, 1.60 or later will have this fix. gRPC ObjC and PHP are not affected by this CVE because they do not support the server feature that has the vulnerability. On Tuesday, October 24, 2023 at 6:56:22 

[grpc-io] Re: grpc libs c++ static linking

>From the error message, linking to abseil library didn't seem to be done correctly. This example (https://github.com/grpc/grpc/blob/master/examples/cpp/helloworld/CMakeLists.txt) would be helpful to understand how it needs to be done. On Thursday, October 5, 2023 at 6:59:17 AM UTC-7 smita

[grpc-io] gRPC-Core Release 1.58.0

This is 1.58.0 (goku ) release announcement for gRPC-Core and the wrapped languages C++, C#, Objective-C, Python, PHP and Ruby. Latest release notes are here . This release

[grpc-io] Re: GRPC latest release v1.56.0 - compile c++/rhel8 (gcc8.5.0) w/ cmake ; to create libprotobuf.so.** together with libgrpc.so.**

Hey Bu, You should be able to see generated protobuf so file with DgRPC_PROTOBUF_PROVIDER=module. I did it myself with the HEAD version of gRPC. BTW, this is only to make gRPC development easy and not for package management. In general, using protobuf repo to generate protobuf so files is

[grpc-io] Re: Building C++ shared library errors

Hey Viral, If you can, it'd be useful to use the latest version of gRPC instead of 1.35. And I tried to read the attached log file but it seems to be truncated. I couldn't find any error messages from it. Would you double-check the file? Regards, Esun. On Saturday, December 3, 2022 at

[grpc-io] Re: c++ gRPC library compiled size - reasonable opportunities for reduction

Hey Kyle, gRPC C++ doesn't have many knobs to reduce the footprint. One option you can try is "--define=grpc_no_xds=true" for bazel. Unfortunately, this option isn't available for cmake but you can take a look at what can be removed by this option since you've made local changes already.

[grpc-io] Re: GRPC Asynchronous Streaming Write() Op in C++

Your understanding is correct but this doesn't necessarily mean that it increases unnecessary latency. gRPC does this for flow control, otherwise the receiver will be overflown by excessive messages. Another thing you can try is to establish multiple connections because some environment can

[grpc-io] gRPC C++ Patch Release for CVE-2023-0286

This patch release is to address [CVE-2023-0286](https://www.openssl.org/news/secadv/20230207.txt). If you meet ALL following conditions, there could be security risk so please consider upgrading gRPC C++. - You're using gRPC C++ - You're using BoringSSL - You're using CRL

[grpc-io] Re: Does GRPC C++ support OpenSSL3.0 compilation?

I don't think gRPC supports OpenSSL 3.0 yet as we have https://github.com/grpc/grpc/pull/31256 open. But it's puzzling to me that you saw many undefined errors. I expected that it should be compile errors due to API breaking changes or runtime errors due to potential API behavior changes. On

[grpc-io] Re: Symmetric crypto used by ALTS record layer

Hi Steven, I'm not an expert on this matter but Google published the ALTS whitepaper; https://cloud.google.com/static/docs/security/encryption-in-transit/application-layer-transport-security/resources/alts-whitepaper.pdf which might answer your question. Regards, Esun. On Friday, August 26,

[grpc-io] Re: Is there a way to install gRPC via CMake without requiring sudo?

I think this should work. If you see this problem only with zlib, you can consider installing gRPC without installing zlib by passing `-DgRPC_ZLIB_PROVIDER=package` when running cmake which means that it will use installed zlib. On Tuesday, December 13, 2022 at 2:26:04 AM UTC-8

[grpc-io] gRFC L102: New gRPC C++ version macros

I've written a proposal to add new version macros to gRPC C++; https://github.com/grpc/proposal/pull/327 -- You received this message because you are subscribed to the Google Groups "grpc.io" group. To unsubscribe from this group and stop receiving emails from it, send an email to

[grpc-io] Re: Building grpc 1.46.3

Hey Oleg, I think this is not because you're using C++17 but because your build might mixed up C++11 (or 14) and C++17 when building all components that gRPC needs. Internally gRPC is being built with C++17 so it shouldn't have a build issue. If you think there is something gRPC needs to fix

[grpc-io] Re: Multi-thread async gRPC server, CPU usage is keeping 20x%

> however, I use the ghz(https://ghz.sh/docs/usage) as the performance tool test the multi-thread async server, the CPU useage between 100%~210%. why? Is there any configure? I don't understand your question. Since your code looks like to use many threads, it could go over 100% implying that

[grpc-io] Re: Running multiple grpc server in the same process

Running multiple gRPC server in the same process should work. If you have core dump, it's more likely there are other problems. On Friday, April 9, 2021 at 11:11:17 AM UTC-7 zhan...@gmail.com wrote: > Hi, > > We have an grpc c++ application which running multiple grpc servers, and > we often

[grpc-io] Re: Detect stream closed from client?

No, you need to call write to notice if it's closed. On Thursday, September 17, 2020 at 9:02:47 PM UTC-7 natalie...@gmail.com wrote: > Hi all, > > I have a GRPC streaming API in C++ that streams from the client to the > server. > > rpc TransferResultChunk(stream TransferResultChunkRequest)

[grpc-io] Re: gRPC slower as iPerf benchmark

Hi Kevin, performance depends on how you organize your client and backend. https://grpc.io/docs/guides/performance/ would be a good starting point to get better performance. Also gRPC has a flow control starting from small size of window so it usually takes some time to get the maximum

[grpc-io] Re: Linking imported proto file to headers in CMakeLists.txt

If you have two proto files, you may want to have the things done in the example twice; one for main.proto and the other for commonutil.proto. And others are pretty regular cmake wiring; You can have a single proto target including two generated sources or two proto target having a dependency.

[grpc-io] Re: hpack_parser.h compiler error

gRPC Core needs C++ to build it. On Monday, January 17, 2022 at 7:15:15 AM UTC-8 Fridolin Siegmund wrote: > Hi! > I want to integrate a grpc client in an existing c project. As an > standalone example (in C!) it compiles and works fine (cmake with > FetchConent to fetch grpc [ >

[grpc-io] Re: Proper design for long RPC calls (C++)

I don't think there is a right way for this. If the client doesn't need to know the result of long-running job, gRPC part would be simple. A server just needs to manage its work; it could be just spawning a dedicated thread handling that job. If the client needs the result, it's just a normal

[grpc-io] Re: Finding out why grpc async service got a longer latency in client side than server

Trace isn't that helpful to understand what happened if you don't know gRPC's internal. Maybe grpc_trace=http gives you some clue as to whether it's a network issue or not. For the latency of the client, I don't understand. Server code looks good and since it's working well without heavy

[grpc-io] Re: Default Recator for Streaming Request

Could you elaborate your suggestion? I cannot see why it's useful. Usually streaming request reactor needs some state to handle writing a series of messages to the peer. On Thursday, March 3, 2022 at 10:02:30 AM UTC-8 Brent Edwards wrote: > This is for a service with callback processing

[grpc-io] Re: Serialized data access for C++ async server

AFAIK there is no way for that. You may have a way by using SerializationTraits that handles (de)serialization but it's not recommended since it's not designed for that purpose. On Saturday, May 1, 2021 at 2:44:49 PM UTC-7 abo...@gmail.com wrote: > Hi, > I have a C++ based gRPC based

[grpc-io] Re: Memory Leakage in gRPC Server

Would you give a small repro we can run and the output of full leak report to know what's leaked? On Wednesday, September 22, 2021 at 6:57:26 PM UTC-7 AmirM wrote: > Hi > > I wrapped a c++ gRPC server in a docker, and I observed when every time I > ran the server docker, memory usage would

[grpc-io] Re: C++ grpc 1.0.0 client connect to grpc 1.38 server got the erro_code 14 without error_message

14 means UNAVAILABLE indicating that the backend isn't available. If you can get extra logging & tracing by https://github.com/grpc/grpc/blob/master/TROUBLESHOOTING.md, it would be helpful for you to understand what happened. And gRPC C++ 1.0 is really old so I recommend to upgrade to the

[grpc-io] Re: memory leak in grpc

I see a couple of new operation without free in your code. So you may want to clean your code first and use MSAN or Valgrind to track down which objects are not freed. If you think that gRPC has its own memory leak, please give us a small repro we can run so that we can investigate it further.

[grpc-io] Re: what is Grpc c++ Release version matching to minimum gcc 4.8.1

Hey Rajesh, I believe that 1.45.2 would be the last version to support GCC 4.8 since the test with gcc 4.8 was removed from 1.46.0. If it isn't working, you can try older versions from 1.44.x. I bet you can find the right one soon. In addition to this, I'd recommend you to stick with the

[grpc-io] gRFC L98: Requiring C++14 in gRPC Core/C++ Library

This is the discussion thread for https://github.com/grpc/proposal/pull/299 Abstract: gRPC starts requiring C++14. -- You received this message because you are subscribed to the Google Groups "grpc.io" group. To unsubscribe from this group and stop receiving emails from it, send an email to

[grpc-io] Re: absl related Compilation Error of grpc cpp serve and client

Would you elaborate a bit about how to build it? If you're using cmake to build gRPC and your application, you shouldn't need to specify linker options. Please checkout https://grpc.io/docs/languages/cpp/quickstart/ too. On Wednesday, January 19, 2022 at 2:27:04 AM UTC-8 cansu gençoğlu wrote: >

Re: [grpc-io] Re: c++: gRPC shutdown() hangs for ever

Hi Ahn, Thanks for the stack trace but I'm a bit confused. Is this what you got while your program was stuck waiting forever? It looks a bit off because Valgrind complains about "Conditional jump or move depends on uninitialised value" and callstack appears to show some deadline inside. On

[grpc-io] gRPC C++ will drop gcc 4.9 support from 1.45

gRPC C++ will change the minimum supported version of gcc from 1.45. Currently it is 4.9+ but it will change to 5.1+. This is decided to be in line with the dependencies of gRPC such as Abseil and GoogleTest which require gcc 5+. From 1.45, the supported compilers of gRPC C++ will be - GCC

[grpc-io] Re: c++: gRPC shutdown() hangs for ever

Would you capture the call-stack when Shutdown function appears to hang? On Wednesday, February 2, 2022 at 10:24:14 PM UTC-8 Anh Tran wrote: > The last commit on my local copy of gRPC is 06 Jan 2022, so not that old. > On Thursday, February 3, 2022 at 4:51:16 AM UTC+10 Easwar Swaminathan >

[grpc-io] gRPC-Core Release 1.43.0

This is the 1.43.0 (green) release announcement for gRPC-Core and the wrapped languages C++, C#, Objective-C, Python, PHP and Ruby. Latest release notes are here . This release contains refinements, improvements, and bug fixes, with highlights

[grpc-io] Re: gRFC L79: C++ API changes on ByteBuffer and Slice

Thanks AJ, I didn't notice the the link actually was linked to the different URL. On Monday, April 19, 2021 at 5:03:50 PM UTC-7 AJ Heller wrote: > Corrected link: https://github.com/grpc/proposal/pull/232. (previous > link pointed to #215) > On Monday, April 19, 2021 at 4:24:39 PM UTC-7

[grpc-io] gRFC L79: C++ API changes on ByteBuffer and Slice

Hi all, I've created a gRFC for C++ API changes on ByteBuffer and Slice. The proposal is here: https://github.com/grpc/proposal/pull/232 Please feel free to comment. Regards, Esun. -- You received this message because you are subscribed to the

[grpc-io] Re: How to explicitly close connection to clients on the server side?

I think there is no way doing this in gRPC C++ so this should be done by an application-side bookkeeping for connections. On Saturday, September 19, 2020 at 4:25:48 PM UTC-7 kishan patel wrote: > I'm using the c++ version of the sever to test my application and I wanted > to know whether

[grpc-io] Re: avoid memory copy in bidi stream

This is more related to the way to use protobuf message efficiently. One thing you can try is to use `release_data()` which gives you an ownership of the data so that you can reuse it. Note that this doesn't necessary guarantee it won't involve any memory copy because it works differently

[grpc-io] Re: Requesting more documentation on "cannot set inq" error

"cannot set inq fd=210 errno=92" isn't an error message but a debug message. setsockopt in the code may or may not succeed depending on the underlying kernel support. Presumably it isn't supported when running in a Kubernetes cluster. On Monday, March 2, 2020 at 7:56:05 AM UTC-8

[grpc-io] Re: Adding service to GRPC Server (CPP)

AFAIK, services cannot be modified once the server instance is built from the builder. On Monday, November 9, 2020 at 1:06:45 PM UTC-8 garog...@gmail.com wrote: > Newbie question: > > Background: Split the server from the actual services, and be able to > add/remove a service from the server

[grpc-io] Re: Check for successful connection?

Sorry for the late answer. As you can said, it can be checked by sending a simple test rpc. Or you can do by monitoring the state of channel via WaitForStateChange method after making the channel start to connect by calling channel->GetState(true) . On Thursday, October 22, 2020 at 12:27:37 PM

[grpc-io] Re: SIGSEGV in ev_epollex_linux.cc

Hi Bryce, Can you try to reproduce this problem with the latest gRPC? Chances are that newer versions of gRPC already fixed this if it's coming from gRPC itself. Regards, Esun. On Friday, October 16, 2020 at 10:02:55 AM UTC-7 Bryce Edwards wrote: > Hi All, > > We're observing a SIGSEGV in