file:// naming service takes a comment after the address as the tag
Maintainers usually reply within 2 days
Assessment
- Difficulty
- 2/5
- Estimated time
- 1-3 hours
- Newbie friendliness
- 35/100
- Issue type
- Bug
- Clarity
- Clearly specified
- Activity status
- Stale
- Tech stack
- cpp
- Domain
- backend-api-design
Research direction
Start with SplitIntoServerAndTag() in src/brpc/policy/file_naming_service.cpp and the related unit test mentioned in the issue. Check how comment parsing handles lines with no tag, then verify that comments are ignored and the server tag is empty. An open pull request, #3735, already contains a fix for this issue.
Written by the indexing model from the issue text.
Description
Describe the bug
In a file:// naming service file, a line with an address and a comment but no tag gets the comment as its tag:
10.24.234.18:8080 # a comment without tag
gives a ServerNode with tag # a comment without tag. docs/en/client.md says "contents after # are comments and ignored", and a comment after a tag (10.24.234.17:8080 tag1 # a comment) is cut correctly.
Because a server is identified by address + tag, the same address with two different comments becomes two instances. The comment is also what NamingServiceFilter and the tag-aware load balancers see.
The cause is in SplitIntoServerAndTag() (src/brpc/policy/file_naming_service.cpp:49-53 at 05abb1c3): the first character after the blanks is taken as the first tag character, and the search for # starts at the next one. list:// and remote file lists use the same function.
To Reproduce
#include <stdio.h>
#include <unistd.h>
#include <string>
#include <vector>
#include "butil/endpoint.h"
#include "butil/files/temp_file.h"
#include "brpc/channel.h"
#include "brpc/naming_service_filter.h"
class RecordingFilter : public brpc::NamingServiceFilter {
public:
bool Accept(const brpc::ServerNode& server) const override {
servers.push_back(server);
return true;
}
mutable std::vector<brpc::ServerNode> servers;
};
int main() {
butil::TempFile tmp;
FILE* fp = fopen(tmp.fname(), "w");
fputs("# This line is ignored\n"
"10.24.234.17:8080 tag1 # a comment\n"
"10.24.234.18:8080 # a comment without tag\n"
"10.24.234.18:8080 # same server, another comment\n"
"10.24.234.19:8080 #\n", fp);
fclose(fp);
RecordingFilter filter;
brpc::ChannelOptions options;
options.ns_filter = &filter;
brpc::Channel channel;
const std::string url = std::string("file://") + tmp.fname();
if (channel.Init(url.c_str(), "rr", &options) != 0) {
return 2;
}
for (size_t i = 0; i < filter.servers.size(); ++i) {
printf("server[%zu] addr=%s tag=\"%s\"\n", i,
butil::endpoint2str(filter.servers[i].addr).c_str(),
filter.servers[i].tag.c_str());
}
fflush(stdout);
_exit(0);
}
Output on master:
server[0] addr=10.24.234.17:8080 tag="tag1"
server[1] addr=10.24.234.18:8080 tag="# a comment without tag"
server[2] addr=10.24.234.18:8080 tag="# same server, another comment"
server[3] addr=10.24.234.19:8080 tag="#"
Expected behavior
server[0] addr=10.24.234.17:8080 tag="tag1"
server[1] addr=10.24.234.18:8080 tag=""
server[2] addr=10.24.234.19:8080 tag=""
Versions
OS: Ubuntu 24.04, x86_64
Compiler: gcc 13.3.0
brpc: master @ 05abb1c3
protobuf: 28.3
Additional context/screenshots
- A related case that this report does not cover: in
10.0.0.3:80#c(no blank before#) the whole token is taken as the address, so by the code the line is rejected as "Invalid address" (read from the code, not run). Should the address also end at#? That could changeunix:paths that contain#. - A PR with the fix (one condition in
SplitIntoServerAndTag()) and a unit test follows. - This report was prepared with an AI coding assistant (Claude Code, model claude-fable-5-2): the bug was found by reading the code and reproduced with the program above on master @ 05abb1c3 on a Linux x86_64 machine; the output shown is from that run.
- Dominant language
- C++
- Stars
- 17.6k
- Forks
- 4.2k
- Avg merge
- 3d 17h
- Merged PRs (30d)
- 72
Getting set up
- Ships a Dockerfile or Docker Compose file
- Has a pull request template
- Read the contributing guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from apache/brpc
-
Difficulty 1/5 1-3 hours Newbie friendliness 88/100
Maintainers usually reply within 2 days
-
endpoint2hostname() prints a byte-swapped port for IPv6 endpointsPossibly taken @0z5a claimed this 8 days ago. Open
Difficulty 2/5 1-3 hours Newbie friendliness 86/100
Maintainers usually reply within 2 days
-
perf(butil): ARM64 (aarch64) performance and modernisation improvements across atomics, CRC32C, spin-wait hints, and fast timersPossibly taken @jimmystewpot claimed this 2 days ago. Open
Difficulty 5/5 Over a week Newbie friendliness 12/100
Maintainers usually reply within 2 days
-
Difficulty 5/5 Over a week Newbie friendliness 25/100
Maintainers usually reply within 2 days
-
enhancement
Difficulty 5/5 Over a week Newbie friendliness 20/100
apache/brpc#3707 · 3 comments ·
Maintainers usually reply within 2 days
Similar issues
-
Difficulty 2/5 1-3 hours Newbie friendliness 77/100
protocolbuffers/protobuf#31768 ·
Maintainers usually reply within 1 day
-
Needs: Triage :mag: Platform: Android
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
react/react-native#60167 ·
Maintainers usually reply within 2 days
-
documentation
Difficulty 2/5 1-3 hours Newbie friendliness 63/100
-
tests: ofx_render_test doesn't compile against FFmpeg 4.x (const AVCodec* to av_find_best_stream)Open
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
NatronGitHub/openfx-io#221 · 1 comment ·
Maintainers usually reply within 1 day
-
bug
Difficulty 2/5 1-3 hours Newbie friendliness 72/100
subsurface/subsurface#5145 ·
Maintainers usually reply within 2 days