add email_exception() to all except: statements.

[monitor.git] / bootman.py

diff --git a/bootman.py b/bootman.py
index 87d8b71..f3ecf72 100755 (executable)
--- a/bootman.py
+++ b/bootman.py
@@ -24,7 +24,7 @@ from unified_model import *
 from emailTxt import mailtxt
 from nodeconfig import network_config_to_str
 import traceback
-import monitorconfig
+import config
 
 import signal
 class Sopen(subprocess.Popen):
@@ -207,7 +207,7 @@ class PlanetLabSession:
                args['port'] = self.port
                args['user'] = 'root'
                args['hostname'] = self.node
-               args['monitordir'] = monitorconfig.MONITOR_SCRIPT_ROOT
+               args['monitordir'] = config.MONITOR_SCRIPT_ROOT
                ssh_port = 22
 
                if self.nosetup:
@@ -324,7 +324,8 @@ def reboot(hostname, config=None, forced_action=None):
                                                        mailtxt.newbootcd_one[1] % args, True, db='bootcd_persistmessages')
 
                loginbase = plc.siteId(hostname)
-               m.send([const.PIEMAIL % loginbase, const.TECHEMAIL % loginbase])
+               emails = plc.getTechEmails(loginbase)
+               m.send(emails) 
 
                print "\tDisabling %s due to out-of-date BOOTCD" % hostname
                api.UpdateNode(hostname, {'boot_state' : 'disable'})
@@ -337,6 +338,8 @@ def reboot(hostname, config=None, forced_action=None):
        try:
                k = SSHKnownHosts(); k.update(node); k.write(); del k
        except:
+               from nodecommon import email_exception
+               email_exception()
                print traceback.print_exc()
                return False
 
@@ -346,8 +349,11 @@ def reboot(hostname, config=None, forced_action=None):
                else:
                        session = PlanetLabSession(node, config.nosetup, config.verbose)
        except Exception, e:
-               print "ERROR setting up session for %s" % hostname
+               msg = "ERROR setting up session for %s" % hostname
+               print msg
                print traceback.print_exc()
+               from nodecommon import email_exception
+               email_exception(msg)
                print e
                return False
 
@@ -361,6 +367,8 @@ def reboot(hostname, config=None, forced_action=None):
                        conn = session.get_connection(config)
                except:
                        print traceback.print_exc()
+                       from nodecommon import email_exception
+                       email_exception()
                        return False
 
        if forced_action == "reboot":
@@ -455,7 +463,8 @@ def reboot(hostname, config=None, forced_action=None):
                                                                                 mailtxt.baddisk[1] % args, True, db='hardware_persistmessages')
 
                        loginbase = plc.siteId(hostname)
-                       m.send([const.PIEMAIL % loginbase, const.TECHEMAIL % loginbase])
+                       emails = plc.getTechEmails(loginbase)
+                       m.send(emails) 
                        conn.set_nodestate('disable')
                        return False
 
@@ -514,6 +523,7 @@ def reboot(hostname, config=None, forced_action=None):
                        ('hardwarerequirefail' , 'Hardware requirements not met'),
                        ('mkfsfail'         , 'while running: Running mkfs.ext2 -q  -m 0 -j /dev/planetlab/vservers failed'),
                        ('nofilereference', "No such file or directory: '/tmp/mnt/sysimg//vservers/.vref/planetlab-f8-i386/etc/hosts'"),
+                       ('kernelcopyfail', "cp: cannot stat `/tmp/mnt/sysimg/boot/kernel-boot': No such file or directory"),
                        ('chrootfail'   , 'Running chroot /tmp/mnt/sysimg'),
                        ('modulefail'   , 'Unable to get list of system modules'),
                        ('writeerror'   , 'write error: No space left on device'),
@@ -541,11 +551,11 @@ def reboot(hostname, config=None, forced_action=None):
        #  By using the sequence identifier, we guarantee that there will be no
        #  frequent loops.  I'm guessing there is a better way to track loops,
        #  though.
-       if not config.force and pflags.getRecentFlag(s):
-               pflags.setRecentFlag(s)
-               pflags.save() 
-               print "... flag is set or it has already run recently. Skipping %s" % node
-               return True
+       #if not config.force and pflags.getRecentFlag(s):
+       #       pflags.setRecentFlag(s)
+       #       pflags.save() 
+       #       print "... flag is set or it has already run recently. Skipping %s" % node
+       #       return True
 
        sequences = {}
 
@@ -583,7 +593,12 @@ def reboot(hostname, config=None, forced_action=None):
                        "bminit-cfg-auth-getplc-update-installinit-validate-rebuildinitrd-netcfg-update3-implementerror-nofilereference-update-debug-done",
                        "bminit-cfg-auth-getplc-update-hardware-installinit-installdisk-exception-mkfsfail-update-debug-done",
                        "bminit-cfg-auth-getplc-installinit-validate-rebuildinitrd-exception-chrootfail-update-debug-done",
+                       "bminit-cfg-auth-getplc-update-hardware-installinit-installdisk-installbootfs-installcfg-installstop-update-installinit-validate-rebuildinitrd-netcfg-disk-update4-update3-update3-kernelcopyfail-exception-update-debug-done",
+                       "bminit-cfg-auth-getplc-hardware-installinit-installdisk-installbootfs-installcfg-installstop-update-installinit-validate-rebuildinitrd-netcfg-disk-update4-update3-update3-kernelcopyfail-exception-update-debug-done",
                        "bminit-cfg-auth-getplc-installinit-validate-exception-noinstall-update-debug-done",
+                       # actual solution appears to involve removing the bad files, and
+                       # continually trying to boot the node.
+                       "bminit-cfg-auth-getplc-update-installinit-validate-rebuildinitrd-netcfg-disk-update4-update3-update3-implementerror-update-debug-done",
                        ]:
                sequences.update({n : "restart_bootmanager_rins"})
 
@@ -612,6 +627,7 @@ def reboot(hostname, config=None, forced_action=None):
                         "bminit-cfg-auth-implementerror-bootcheckfail-update-implementerror-bootupdatefail-done",
                         "bminit-cfg-auth-getplc-update-installinit-validate-rebuildinitrd-netcfg-update3-implementerror-nospace-update-debug-done",
                         "bminit-cfg-auth-getplc-hardware-installinit-installdisk-installbootfs-exception-downloadfail-update-debug-done",
+                        "bminit-cfg-auth-getplc-update-installinit-validate-implementerror-update-debug-done",
                         ]:
                sequences.update({n: "restart_node_boot"})
 
@@ -668,7 +684,7 @@ def reboot(hostname, config=None, forced_action=None):
                m = PersistMessage(hostname, mailtxt.unknownsequence[0] % args,
                                                                         mailtxt.unknownsequence[1] % args, False, db='unknown_persistmessages')
                m.reset()
-               m.send(['monitor-list@lists.planet-lab.org'])
+               m.send([config.cc_email]) 
 
                conn.restart_bootmanager('boot')
 
@@ -706,7 +722,7 @@ def reboot(hostname, config=None, forced_action=None):
                        m = PersistMessage(hostname, "Suspicous error from BootManager on %s" % args,
                                                                                 mailtxt.unknownsequence[1] % args, False, db='suspect_persistmessages')
                        m.reset()
-                       m.send(['monitor-list@lists.planet-lab.org'])
+                       m.send([config.cc_email]) 
 
                        conn.restart_bootmanager('boot')
 
@@ -717,7 +733,8 @@ def reboot(hostname, config=None, forced_action=None):
                        m = PersistMessage(hostname,  mailtxt.plnode_cfg[0] % args,  mailtxt.plnode_cfg[1] % args, 
                                                                True, db='nodeid_persistmessages')
                        loginbase = plc.siteId(hostname)
-                       m.send([const.PIEMAIL % loginbase, const.TECHEMAIL % loginbase])
+                       emails = plc.getTechEmails(loginbase)
+                       m.send(emails) 
                        conn.dump_plconf_file()
                        conn.set_nodestate('disable')
 
@@ -726,10 +743,11 @@ def reboot(hostname, config=None, forced_action=None):
                        args = {}
                        args['hostname'] = hostname
                        args['bmlog'] = conn.get_bootmanager_log().read()
-                       m = PersistMessage(hostname,  mailtxt.plnode_network[0] % args,  mailtxt.plnode_cfg[1] % args, 
+                       m = PersistMessage(hostname,  mailtxt.plnode_cfg[0] % args,  mailtxt.plnode_cfg[1] % args, 
                                                                True, db='nodenet_persistmessages')
                        loginbase = plc.siteId(hostname)
-                       m.send([const.PIEMAIL % loginbase, const.TECHEMAIL % loginbase])
+                       emails = plc.getTechEmails(loginbase)
+                       m.send(emails) 
                        conn.dump_plconf_file()
                        conn.set_nodestate('disable')
 
@@ -744,7 +762,8 @@ def reboot(hostname, config=None, forced_action=None):
                                                                mailtxt.newalphacd_one[1] % args, True, db='bootcd_persistmessages')
 
                        loginbase = plc.siteId(hostname)
-                       m.send([const.PIEMAIL % loginbase, const.TECHEMAIL % loginbase])
+                       emails = plc.getTechEmails(loginbase)
+                       m.send(emails) 
 
                        print "\tDisabling %s due to out-of-date BOOTCD" % hostname
                        conn.set_nodestate('disable')
@@ -762,7 +781,8 @@ def reboot(hostname, config=None, forced_action=None):
                                                                                 mailtxt.baddisk[1] % args, True, db='hardware_persistmessages')
 
                        loginbase = plc.siteId(hostname)
-                       m.send([const.PIEMAIL % loginbase, const.TECHEMAIL % loginbase])
+                       emails = plc.getTechEmails(loginbase)
+                       m.send(emails) 
                        conn.set_nodestate('disable')
 
                elif sequences[s] == "update_hardware_email":
@@ -774,7 +794,8 @@ def reboot(hostname, config=None, forced_action=None):
                                                                                 mailtxt.minimalhardware[1] % args, True, db='minhardware_persistmessages')
 
                        loginbase = plc.siteId(hostname)
-                       m.send([const.PIEMAIL % loginbase, const.TECHEMAIL % loginbase])
+                       emails = plc.getTechEmails(loginbase)
+                       m.send(emails) 
                        conn.set_nodestate('disable')
 
                elif sequences[s] == "bad_dns_email":
@@ -784,6 +805,8 @@ def reboot(hostname, config=None, forced_action=None):
                                node = api.GetNodes(hostname)[0]
                                net = api.GetNodeNetworks(node['nodenetwork_ids'])[0]
                        except:
+                               from nodecommon import email_exception
+                               email_exception()
                                print traceback.print_exc()
                                # TODO: api error. skip email, b/c all info is not available,
                                # flag_set will not be recorded.
@@ -797,7 +820,8 @@ def reboot(hostname, config=None, forced_action=None):
                                                                                 mailtxt.baddns[1] % args, True, db='baddns_persistmessages')
 
                        loginbase = plc.siteId(hostname)
-                       m.send([const.PIEMAIL % loginbase, const.TECHEMAIL % loginbase])
+                       emails = plc.getTechEmails(loginbase)
+                       m.send(emails) 
                        conn.set_nodestate('disable')
 
        if flag_set: