Apache
/
hadoop
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

YARN-8587. Added retries for fetching docker exit code. 

Contributed by Charo Zhang

(cherry picked from commit c16c49b8c3)
This commit is contained in:
Eric Yang 2018-10-24 17:28:23 -04:00
parent 4a64dab0dd
commit ac85aa80d9
1 changed files with 17 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

32
hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/main/native/container-executor/impl/container-executor.c
View File

@ -1765,20 +1765,22 @@ int launch_docker_container_as_user(const char * user, const char *app_id,
docker_binary, container_id);
fprintf(LOGFILE, "Obtaining the exit code...\n");
fprintf(LOGFILE, "Docker inspect command: %s\n", docker_inspect_exitcode_command);
FILE* inspect_exitcode_docker = popen(docker_inspect_exitcode_command, "r");
if(inspect_exitcode_docker == NULL) {
fprintf(ERRORFILE, "Done with inspect_exitcode, inspect_exitcode_docker is null\n");
fflush(ERRORFILE);
exit_code = -1;
goto cleanup;
}
res = fscanf (inspect_exitcode_docker, "%d", &exit_code);
if (pclose (inspect_exitcode_docker) != 0 || res <= 0) {
fprintf (ERRORFILE,
"Could not inspect docker to get exitcode: %s.\n", docker_inspect_exitcode_command);
fflush(ERRORFILE);
exit_code = -1;
goto cleanup;
int count = 0;
int max_retries = get_max_retries(&CFG);
while (count < max_retries) {
FILE* inspect_exitcode_docker = popen(docker_inspect_exitcode_command, "r");
res = fscanf (inspect_exitcode_docker, "%d", &exit_code);
if (pclose (inspect_exitcode_docker) != 0 || res <= 0) {
fprintf (ERRORFILE, "Could not inspect docker to get Exit code %s.\n", docker_inspect_exitcode_command);
fflush(ERRORFILE);
exit_code = -1;
} else {
if (exit_code != 0) {
break;
}
}
sleep(3);
count++;
}
fprintf(LOGFILE, "Exit code from docker inspect: %d\n", exit_code);
@ -2708,4 +2710,4 @@ int remove_docker_container(char**argv, int argc) {
exit_code = clean_docker_cgroups(yarn_hierarchy, container_id);
}
return exit_code;
}
}