Robustness work for port_server startup
diff --git a/tools/run_tests/port_server.py b/tools/run_tests/port_server.py
index b953df9..4e473af 100755
--- a/tools/run_tests/port_server.py
+++ b/tools/run_tests/port_server.py
@@ -42,7 +42,7 @@
# increment this number whenever making a change to ensure that
# the changes are picked up by running CI servers
# note that all changes must be backwards compatible
-_MY_VERSION = 2
+_MY_VERSION = 4
if len(sys.argv) == 2 and sys.argv[1] == 'dump_version':
@@ -52,8 +52,13 @@
argp = argparse.ArgumentParser(description='Server for httpcli_test')
argp.add_argument('-p', '--port', default=12345, type=int)
+argp.add_argument('-l', '--logfile', default=None, type=str)
args = argp.parse_args()
+if args.logfile is not None:
+ sys.stderr = open(args.logfile, 'w')
+ sys.stdout = sys.stderr
+
print 'port server running on port %d' % args.port
pool = []
@@ -146,6 +151,6 @@
httpd = BaseHTTPServer.HTTPServer(('', args.port), Handler)
while keep_running:
httpd.handle_request()
+ sys.stderr.flush()
print 'done'
-
diff --git a/tools/run_tests/run_tests.py b/tools/run_tests/run_tests.py
index 048ab90..7137759 100755
--- a/tools/run_tests/run_tests.py
+++ b/tools/run_tests/run_tests.py
@@ -43,6 +43,8 @@
import socket
import subprocess
import sys
+import tempfile
+import traceback
import time
import xml.etree.cElementTree as ET
import urllib2
@@ -704,35 +706,50 @@
urllib2.urlopen('http://localhost:%d/quitquitquit' % port_server_port).read()
time.sleep(1)
if not running:
- print 'starting port_server'
- port_log = open('portlog.txt', 'w')
+ fd, logfile = tempfile.mkstemp()
+ os.close(fd)
+ print 'starting port_server, with log file %s' % logfile
port_server = subprocess.Popen(
- [sys.executable, 'tools/run_tests/port_server.py', '-p', '%d' % port_server_port],
- stderr=subprocess.STDOUT,
- stdout=port_log)
+ [sys.executable, 'tools/run_tests/port_server.py', '-p', '%d' % port_server_port, '-l', logfile],
+ close_fds=True)
+ time.sleep(1)
# ensure port server is up
waits = 0
while True:
if waits > 10:
+ print 'killing port server due to excessive start up waits'
port_server.kill()
if port_server.poll() is not None:
print 'port_server failed to start'
- port_log = open('portlog.txt', 'r').read()
- print port_log
- sys.exit(1)
+ # try one final time: maybe another build managed to start one
+ time.sleep(1)
+ try:
+ urllib2.urlopen('http://localhost:%d/get' % port_server_port,
+ timeout=1).read()
+ print 'last ditch attempt to contact port server succeeded'
+ break
+ except:
+ traceback.print_exc();
+ port_log = open(logfile, 'r').read()
+ print port_log
+ sys.exit(1)
try:
urllib2.urlopen('http://localhost:%d/get' % port_server_port,
timeout=1).read()
+ print 'port server is up and ready'
break
except socket.timeout:
print 'waiting for port_server: timeout'
- time.sleep(0.5)
+ traceback.print_exc();
+ time.sleep(1)
waits += 1
except urllib2.URLError:
print 'waiting for port_server: urlerror'
- time.sleep(0.5)
+ traceback.print_exc();
+ time.sleep(1)
waits += 1
except:
+ traceback.print_exc();
port_server.kill()
raise