# Attempt to reboot a node in debug state.
import plc
-import auth
-api = plc.PLC(auth.auth, auth.plc)
+api = plc.getAuthAPI()
import sys
import os
import subprocess
import time
-import soltesz
+import database
+import moncommands
from sets import Set
import ssh.pxssh as pxssh
import ssh.pexpect as pexpect
from unified_model import *
from emailTxt import mailtxt
+from nodeconfig import network_config_to_str
+import traceback
+import monitorconfig
import signal
class Sopen(subprocess.Popen):
from Rpyc.Utils import *
def get_fbnode(node):
- fb = soltesz.dbLoad("findbad")
+ fb = database.dbLoad("findbad")
fbnode = fb['nodes'][node]['values']
return fbnode
def dump_plconf_file(self):
c = self.c
- c.modules.sys.path.append("/tmp/source/")
- c.modules.os.chdir('/tmp/source')
+ self.c.modules.sys.path.append("/tmp/source/")
+ self.c.modules.os.chdir('/tmp/source')
log = c.modules.BootManager.log('/tmp/new.log')
bm = c.modules.BootManager.BootManager(log,'boot')
def compare_and_repair_nodekeys(self):
c = self.c
- c.modules.sys.path.append("/tmp/source/")
- c.modules.os.chdir('/tmp/source')
+ self.c.modules.sys.path.append("/tmp/source/")
+ self.c.modules.os.chdir('/tmp/source')
log = c.modules.BootManager.log('/tmp/new.log')
bm = c.modules.BootManager.BootManager(log,'boot')
args['port'] = self.port
args['user'] = 'root'
args['hostname'] = self.node
- args['monitordir'] = "/home/soltesz/monitor"
+ args['monitordir'] = monitorconfig.MONITOR_SCRIPT_ROOT
ssh_port = 22
if self.nosetup:
return
# COPY Rpyc files to host
- cmd = "rsync -qv -az -e ssh %(monitordir)s/Rpyc-2.45-2.3/ %(user)s@%(hostname)s:Rpyc 2> /dev/null" % args
+ cmd = "rsync -qv -az -e ssh %(monitordir)s/Rpyc/ %(user)s@%(hostname)s:Rpyc 2> /dev/null" % args
if self.verbose: print cmd
# TODO: Add timeout
timeout = 120
- localos = soltesz.CMD()
+ localos = moncommands.CMD()
ret = localos.system(cmd, timeout)
print ret
t1 = time.time()
# KILL any already running servers.
- ssh = soltesz.SSH(args['user'], args['hostname'], ssh_port)
+ ssh = moncommands.SSH(args['user'], args['hostname'], ssh_port)
(ov,ev) = ssh.run_noexcept2("""<<\EOF
rm -f out.log
echo "kill server" >> out.log
# TODO: the read() here may block indefinitely. Need a better
# approach therefore, that includes a timeout.
#ret = self.command.stdout.read(5)
- ret = soltesz.read_t(self.command.stdout, 5)
+ ret = moncommands.read_t(self.command.stdout, 5)
t2 = time.time()
if 'READY' in ret:
try:
k = SSHKnownHosts(); k.update(node); k.write(); del k
except:
- import traceback; print traceback.print_exc()
+ print traceback.print_exc()
return False
try:
session = PlanetLabSession(node, config.nosetup, config.verbose)
except Exception, e:
print "ERROR setting up session for %s" % hostname
- import traceback; print traceback.print_exc()
+ print traceback.print_exc()
print e
return False
time.sleep(session.timeout*4)
conn = session.get_connection(config)
except:
- import traceback; print traceback.print_exc()
+ print traceback.print_exc()
return False
('noinstall' , 'notinstalled'),
('bziperror' , 'bzip2: Data integrity error when decompressing.'),
('noblockdev' , "No block devices detected."),
+ ('dnserror' , 'Name or service not known'),
('downloadfail' , 'Unable to download main tarball /boot/bootstrapfs-planetlab-i386.tar.bz2 from server.'),
('disktoosmall' , 'The total usable disk size of all disks is insufficient to be usable as a PlanetLab node.'),
('hardwarerequirefail' , 'Hardware requirements not met'),
for n in ["bminit-cfg-auth-getplc-update-installinit-validate-rebuildinitrd-netcfg-update3-disk-update4-done",
"bminit-cfg-auth-getplc-installinit-validate-rebuildinitrd-netcfg-update3-disk-update4-update3-exception-protoerror-update-protoerror-debug-done",
"bminit-cfg-auth-getplc-installinit-validate-rebuildinitrd-netcfg-disk-update4-update3-update3-implementerror-bootupdatefail-update-debug-done",
+
+ "bminit-cfg-auth-getplc-installinit-validate-rebuildinitrd-netcfg-disk-update4-update3-update3-exception-protoerror-update-protoerror-debug-done",
+
"bminit-cfg-auth-getplc-installinit-validate-rebuildinitrd-netcfg-update3-disk-update4-update3-exception-protoerror-update-debug-done",
"bminit-cfg-auth-getplc-installinit-validate-rebuildinitrd-netcfg-disk-update4-update3-exception-chrootfail-update-debug-done",
"bminit-cfg-auth-getplc-update-debug-done",
"bminit-cfg-auth-protoerror-exception-update-protoerror-debug-done",
"bminit-cfg-auth-protoerror-exception-update-bootupdatefail-authfail-debug-done",
"bminit-cfg-auth-protoerror-exception-update-debug-done",
+ "bminit-cfg-auth-getplc-exception-protoerror-update-debug-done",
"bminit-cfg-auth-getplc-implementerror-update-debug-done",
]:
sequences.update({n : "restart_bootmanager_boot"})
# broken_hardware_email
sequences.update({"bminit-cfg-auth-getplc-update-hardware-exception-hardwarerequirefail-update-debug-done" : "broken_hardware_email"})
+ # bad_dns_email
+ sequences.update({"bminit-cfg-update-implementerror-bootupdatefail-dnserror-update-implementerror-bootupdatefail-dnserror-done" : "bad_dns_email"})
+
flag_set = True
m.send([policy.PIEMAIL % loginbase, policy.TECHEMAIL % loginbase])
conn.set_nodestate('disable')
+ elif sequences[s] == "bad_dns_email":
+ print "...NOTIFYING OWNERS OF DNS FAILURE on %s!!!" % hostname
+ args = {}
+ try:
+ node = api.GetNodes(hostname)[0]
+ net = api.GetNodeNetworks(node['nodenetwork_ids'])[0]
+ except:
+ print traceback.print_exc()
+ # TODO: api error. skip email, b/c all info is not available,
+ # flag_set will not be recorded.
+ return False
+ nodenet_str = network_config_to_str(net)
+
+ args['hostname'] = hostname
+ args['network_config'] = nodenet_str
+ args['nodenetwork_id'] = net['nodenetwork_id']
+ m = PersistMessage(hostname, mailtxt.baddns[0] % args,
+ mailtxt.baddns[1] % args, True, db='baddns_persistmessages')
+
+ loginbase = plc.siteId(hostname)
+ m.send([policy.PIEMAIL % loginbase, policy.TECHEMAIL % loginbase])
+ conn.set_nodestate('disable')
+
if flag_set:
pflags.setRecentFlag(s)
pflags.save()
help="Extra quiet output messages.")
parser.add_option("", "--verbose", dest="verbose", action="store_true",
help="Extra debug output messages.")
+ parser.add_option("", "--nonet", dest="nonet", action="store_true",
+ help="Do not setup the network, use existing log files to re-run a test pass.")
parser.add_option("", "--collect", dest="collect", action="store_true",
help="No action, just collect dmesg, and bm.log")
parser.add_option("", "--nosetup", dest="nosetup", action="store_true",