update installation instructions and support iperf3

README.md

@@ -62,8 +62,8 @@ source $HOME/.local/bin/env
 
 3. Set up the environment:
 ```bash
-sudo apt install iperf -y
-uv venv
+sudo apt install iperf -y # or conda install -c conda-forge iperf
+uv venv --python 3.12.1
 source .venv/bin/activate
 uv sync --extra all
 git submodule update --init --recursive

src/zeroband/comms.py

@@ -29,6 +29,24 @@
 BENCH_TENSOR_SIZE = 1_000_000
 
 
+def _detect_iperf_version() -> str:
+    """Detect which version of iperf is available on the system.
+
+    Returns:
+        str: Either 'iperf3' or 'iperf' depending on which is available.
+        Prefers iperf3 if both are available.
+    """
+    try:
+        subprocess.run(["iperf3", "--version"], capture_output=True, check=True)
+        return "iperf3"
+    except (subprocess.SubprocessError, FileNotFoundError):
+        try:
+            subprocess.run(["iperf", "--version"], capture_output=True, check=True)
+            return "iperf"
+        except (subprocess.SubprocessError, FileNotFoundError):
+            raise RuntimeError("Neither iperf nor iperf3 is available on the system")
+
+
 class ElasticDeviceMesh:
     """A class to manage the process groups for elastic training without restarts.
 
@@ -500,10 +518,11 @@ def _start_iperf_server(self) -> None:
 
             iperf_addr = get_ip_address(IPERF_IFNAME)
             iperf_port = IPERF_PORT + self.world_info.global_rank
-            cmd: List[str] = ["iperf", "-s", "-p", str(iperf_port)]
+            iperf_cmd = _detect_iperf_version()
+            cmd: List[str] = [iperf_cmd, "-s", "-p", str(iperf_port)]
             self.server_process = subprocess.Popen(cmd, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
             self.god_store.set(f"iperf_{self.world_info.global_unique_id}", f"{iperf_addr}:{iperf_port}")
-            self._logger.info(f"Started iperf server on {iperf_addr} with port {iperf_port}")
+            self._logger.info(f"Started {iperf_cmd} server on {iperf_addr} with port {iperf_port}")
         except Exception as e:
             self._logger.error(f"Failed to start iperf server: {str(e)}")
             raise
@@ -529,8 +548,9 @@ def measure_bandwidth(self, target_host: str, target_port: int) -> int:
             int: The time taken to transfer 10Tb of data in seconds
         """
         try:
+            iperf_cmd = _detect_iperf_version()
             cmd: List[str] = [
-                "iperf",
+                iperf_cmd,
                 "-c",
                 target_host,
                 "-p",