class ResourceManagementDB(object):
"""
The ResourceManagementDB class is a front-end to the ResourceManagementDB MySQL db.
If exposes four basic methods:
- insert
- update
- get
- delete
all them defined on the MySQL monkey class.
Moreover, there are a set of key-worded parameters that can be used, specially
on the getX and deleteX functions ( to know more, again, check the MySQL monkey
documentation ).
The DB schema has NO foreign keys, so there may be some small consistency checks,
called validators on the insert and update functions.
The simplest way to instantiate an object of type :class:`ResourceManagementDB`
is simply by calling
>>> rmDB = ResourceManagementDB()
This way, it will use the standard :mod:`DIRAC.Core.Base.DB`.
But there's the possibility to use other DB classes.
For example, we could pass custom DB instantiations to it,
provided the interface is the same exposed by :mod:`DIRAC.Core.Base.DB`.
>>> AnotherDB = AnotherDBClass()
>>> rmDB = ResourceManagementDB( DBin = AnotherDB )
Alternatively, for testing purposes, you could do:
>>> from mock import Mock
>>> mockDB = Mock()
>>> rmDB = ResourceManagementDB( DBin = mockDB )
Or, if you want to work with a local DB, providing it's mySQL:
>>> rmDB = ResourceManagementDB( DBin = [ 'UserName', 'Password' ] )
The ResourceStatusDB also exposes database Schema information, either on a
dictionary or on a MySQLSchema tree object.
- getSchema
- inspectSchema
Alternatively, we can access the MySQLSchema XML and tree as follows:
>>> rmDB = ResourceManagementDB()
>>> xml = rmDB.mm.SCHEMA
>>> tree = rmDB.mm.mSchema
>>> tree
>>> tree.
>>> tree.PolicyResult.Name
"""
def __init__(self, *args, **kwargs):
"""Constructor."""
if len(args) == 1:
if isinstance(args[0], str):
maxQueueSize = 10
if isinstance(args[0], int):
maxQueueSize = args[0]
elif len(args) == 2:
maxQueueSize = args[1]
elif len(args) == 0:
maxQueueSize = 10
if 'DBin' in kwargs.keys():
dbIn = kwargs['DBin']
if isinstance(dbIn, list):
from DIRAC.Core.Utilities.MySQL import MySQL
self.db = MySQL('localhost', dbIn[0], dbIn[1],
'ResourceManagementDB')
else:
self.db = dbIn
else:
from DIRAC.Core.Base.DB import DB
self.db = DB('ResourceManagementDB',
'ResourceStatus/ResourceManagementDB', maxQueueSize)
self.mm = MySQLMonkey(self)
@CheckDBExecution
@ValidateDBTypes
def insert(self, params, meta):
"""
Inserts args in the DB making use of kwargs where parameters such as
the table are specified ( filled automatically by the Client). In order to
do the insertion, it uses MySQLMonkey to do the parsing, execution and
error handling. Typically you will not pass kwargs to this function, unless
you know what are you doing and you have a very special use case.
:Parameters:
**params** - `dict`
arguments for the mysql query ( must match table columns ! ).
**meta** - `dict`
metadata for the mysql query. It must contain, at least, `table` key
with the proper table name.
:return: S_OK() || S_ERROR()
"""
return self.mm.insert(params, meta)
@CheckDBExecution
@ValidateDBTypes
def update(self, params, meta):
"""
Updates row with values given on args. The row selection is done using the
default of MySQLMonkey ( column.primary or column.keyColumn ). It can be
modified using kwargs, but it is not explained here. The table keyword
argument is mandatory, and filled automatically by the Client. Typically
you will not pass kwargs to this function, unless you know what are you
doing and you have a very special use case.
:Parameters:
**params** - `dict`
arguments for the mysql query ( must match table columns ! ).
**meta** - `dict`
metadata for the mysql query. It must contain, at least, `table` key
with the proper table name.
:return: S_OK() || S_ERROR()
"""
return self.mm.update(params, meta)
@CheckDBExecution
@ValidateDBTypes
def get(self, params, meta):
"""
Uses arguments to build conditional SQL statement ( WHERE ... ). If the
sql statement desired is more complex, you can use kwargs to interact with
the MySQLStatement parser and generate a more sophisticated query.
:Parameters:
**params** - `dict`
arguments for the mysql query ( must match table columns ! ).
**meta** - `dict`
metadata for the mysql query. It must contain, at least, `table` key
with the proper table name.
:return: S_OK() || S_ERROR()
"""
return self.mm.get(params, meta)
@CheckDBExecution
@ValidateDBTypes
def delete(self, params, meta):
"""
Uses arguments to build conditional SQL statement ( WHERE ... ). If the
sql statement desired is more complex, you can use kwargs to interact with
the MySQLStatement parser and generate a more sophisticated query. There is
only one forbidden query, with all parameters None ( this would mean a query
of the type `DELETE * from TableName` ). The usage of kwargs is the same
as in the get function.
:Parameters:
**params** - `dict`
arguments for the mysql query ( must match table columns ! ).
**meta** - `dict`
metadata for the mysql query. It must contain, at least, `table` key
with the proper table name.
:return: S_OK() || S_ERROR()
"""
return self.mm.delete(params, meta)
@CheckDBExecution
def getSchema(self):
"""
Returns a dictionary with database schema, this includes table and column
names. It has two variants, columns and keyUsage. The first one has at least,
as many keys as keyUsage, it is the complete schema. The second one is the
one used for the default updates and selects -- not taking into account
auto_increment fields, but taking into account primary and keyUsage fields.
:Parameters:
`None`
:return: S_OK()
"""
return {'OK': True, 'Value': self.mm.SCHEMA}
@CheckDBExecution
def inspectSchema(self):
"""
Returns an object which represents the database schema and can be browsed.
>>> db = ResourceManagementDB()
>>> schema = db.inspectSchema()[ 'Value' ]
>>> schema
Schema 123:
<TableName1>,<TableName2>...
>>> schema.TableName1
Table TableName1:
<ColumnName1>,<ColumnName2>..
Every column has a few attributes ( primary, keyUsage, extra, position,
dataType and charMaxLen ).
:Parameters:
`None`
:return: S_OK()
"""
return {'OK': True, 'Value': self.mm.mSchema}
################################################################################
#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF
################################################################################
## Web functions
################################################################################
#
################################################################################
#
# def getDownTimesWeb(self, selectDict, _sortList = [], startItem = 0, maxItems = 1000):
# """
# Get downtimes registered in the RSS DB (with a web layout)
#
# :params:
# :attr:`selectDict`: { 'Granularity':['Site', 'Resource'], 'Severity': ['OUTAGE', 'AT_RISK']}
#
# :attr:`sortList`
#
# :attr:`startItem`
#
# :attr:`maxItems`
# """
#
# granularity = selectDict['Granularity']
# severity = selectDict['Severity']
#
# if not isinstance(granularity, list):
# granularity = [granularity]
# if not isinstance(severity, list):
# severity = [severity]
#
# paramNames = ['Granularity', 'Name', 'Severity', 'When']
#
# req = "SELECT Granularity, Name, Reason FROM PolicyRes WHERE "
# req = req + "PolicyName LIKE 'DT_%' AND Reason LIKE \'%found%\' "
# req = req + "AND Granularity in (%s)" %(','.join(['"'+x.strip()+'"' for x in granularity]))
# resQuery = self.db._query(req)
# if not resQuery['OK']:
# raise RSSManagementDBException, resQuery['Message']
# if not resQuery['Value']:
# records = []
# else:
# resQuery = resQuery['Value']
# records = []
# for tuple_ in resQuery:
# sev = tuple_[2].split()[2]
# if sev not in severity:
# continue
# when = tuple_[2].split(sev)[1][1:]
# if when == '':
# when = 'Ongoing'
# records.append([tuple_[0], tuple_[1], sev, when])
#
# finalDict = {}
# finalDict['TotalRecords'] = len(records)
# finalDict['ParameterNames'] = paramNames
#
# # Return all the records if maxItems == 0 or the specified number otherwise
# if maxItems:
# finalDict['Records'] = records[startItem:startItem+maxItems]
# else:
# finalDict['Records'] = records
#
# finalDict['Extras'] = None
#
# return finalDict
#
################################################################################
class ResourceStatusDB(object):
"""
The ResourceStatusDB class is a front-end to the ResourceStatusDB MySQL db.
It exposes four basic methods:
- insert
- update
- get
- delete
all them defined on the MySQL monkey class.
Moreover, there are a set of key-worded parameters that can be used, specially
on the getX and deleteX functions ( to know more, again, check the MySQL monkey
documentation ).
The DB schema has NO foreign keys, so there may be some small consistency checks,
called validators on the insert and update functions.
The simplest way to instantiate an object of type :class:`ResourceStatusDB`
is simply by calling
>>> rsDB = ResourceStatusDB()
This way, it will use the standard :mod:`DIRAC.Core.Base.DB`.
But there's the possibility to use other DB classes.
For example, we could pass custom DB instantiations to it,
provided the interface is the same exposed by :mod:`DIRAC.Core.Base.DB`.
>>> AnotherDB = AnotherDBClass()
>>> rsDB = ResourceStatusDB( DBin = AnotherDB )
Alternatively, for testing purposes, you could do:
>>> from mock import Mock
>>> mockDB = Mock()
>>> rsDB = ResourceStatusDB( DBin = mockDB )
Or, if you want to work with a local DB, providing it's mySQL:
>>> rsDB = ResourceStatusDB( DBin = [ 'UserName', 'Password' ] )
The ResourceStatusDB also exposes database Schema information, either on a
dictionary or on a MySQLSchema tree object.
- getSchema
- inspectSchema
Alternatively, we can access the MySQLSchema XML and tree as follows:
>>> rsDB = ResourceStatusDB()
>>> xml = rsDB.mm.SCHEMA
>>> tree = rsDB.mm.mSchema
>>> tree
>>> tree.GridSite
>>> tree.GridSite.GridTier
"""
def __init__(self, *args, **kwargs):
"""Constructor."""
if len(args) == 1:
if isinstance(args[0], str):
maxQueueSize = 10
if isinstance(args[0], int):
maxQueueSize = args[0]
elif len(args) == 2:
maxQueueSize = args[1]
elif len(args) == 0:
maxQueueSize = 10
if 'DBin' in kwargs.keys():
dbIn = kwargs['DBin']
if isinstance(dbIn, list):
from DIRAC.Core.Utilities.MySQL import MySQL
self.db = MySQL('localhost', dbIn[0], dbIn[1],
'ResourceStatusDB')
else:
self.db = dbIn
else:
from DIRAC.Core.Base.DB import DB
self.db = DB('ResourceStatusDB', 'ResourceStatus/ResourceStatusDB',
maxQueueSize)
self.mm = MySQLMonkey(self)
@CheckDBExecution
@ValidateDBTypes
def insert(self, params, meta):
"""
Inserts args in the DB making use of kwargs where parameters such as
the table are specified ( filled automatically by the Client). In order to
do the insertion, it uses MySQLMonkey to do the parsing, execution and
error handling. Typically you will not pass kwargs to this function, unless
you know what are you doing and you have a very special use case.
:Parameters:
**params** - `dict`
arguments for the mysql query ( must match table columns ! ).
**meta** - `dict`
metadata for the mysql query. It must contain, at least, `table` key
with the proper table name.
:return: S_OK() || S_ERROR()
"""
return self.mm.insert(params, meta)
@CheckDBExecution
@ValidateDBTypes
def update(self, params, meta):
"""
Updates row with values given on args. The row selection is done using the
default of MySQLMonkey ( column.primary or column.keyColumn ). It can be
modified using kwargs, but it is not explained here. The table keyword
argument is mandatory, and filled automatically by the Client. Typically
you will not pass kwargs to this function, unless you know what are you
doing and you have a very special use case.
:Parameters:
**params** - `dict`
arguments for the mysql query ( must match table columns ! ).
**meta** - `dict`
metadata for the mysql query. It must contain, at least, `table` key
with the proper table name.
:return: S_OK() || S_ERROR()
"""
return self.mm.update(params, meta)
@CheckDBExecution
@ValidateDBTypes
def get(self, params, meta):
"""
Uses arguments to build conditional SQL statement ( WHERE ... ). If the
sql statement desired is more complex, you can use kwargs to interact with
the MySQLStatement parser and generate a more sophisticated query.
:Parameters:
**params** - `dict`
arguments for the mysql query ( must match table columns ! ).
**meta** - `dict`
metadata for the mysql query. It must contain, at least, `table` key
with the proper table name.
:return: S_OK() || S_ERROR()
"""
return self.mm.get(params, meta)
@CheckDBExecution
@ValidateDBTypes
def delete(self, params, meta):
"""
Uses arguments to build conditional SQL statement ( WHERE ... ). If the
sql statement desired is more complex, you can use kwargs to interact with
the MySQLStatement parser and generate a more sophisticated query. There is
only one forbidden query, with all parameters None ( this would mean a query
of the type `DELETE * from TableName` ). The usage of kwargs is the same as in
the get function.
:Parameters:
**params** - `dict`
arguments for the mysql query ( must match table columns ! ).
**meta** - `dict`
metadata for the mysql query. It must contain, at least, `table` key
with the proper table name.
:return: S_OK() || S_ERROR()
"""
return self.mm.delete(params, meta)
@CheckDBExecution
def getSchema(self):
"""
Returns a dictionary with database schema, this includes table and column
names. It has two variants, columns and keyUsage. The first one has at least,
as many keys as keyUsage, it is the complete schema. The second one is the
one used for the default updates and selects -- not taking into account
auto_increment fields, but taking into account primary and keyUsage fields.
:Parameters:
`None`
:return: S_OK()
"""
return {'OK': True, 'Value': self.mm.SCHEMA}
@CheckDBExecution
def inspectSchema(self):
"""
Returns an object which represents the database schema and can be browsed.
>>> db = ResourceStatusDB()
>>> schema = db.inspectSchema()[ 'Value' ]
>>> schema
Schema 123:
<TableName1>,<TableName2>...
>>> schema.TableName1
Table TableName1:
<ColumnName1>,<ColumnName2>..
Every column has a few attributes ( primary, keyUsage, extra, position,
dataType and charMaxLen ).
:Parameters:
`None`
:return: S_OK()
"""
return {'OK': True, 'Value': self.mm.mSchema}
################################################################
# #
# VALIDATION ?? #
# #
# o Site ->(IU) SiteType #
# o Service ->(IU) ServiceType, Site #
# o Resource ->(IU) ResourceType, ServiceType #
# o StorageElement ->(IU) Resource #
# o GridSite ->(IU) __none__ #
# o *Status ->(IU) *,StatusType,Status #
# o *ScheduledStatus ->(IU) *,StatusType,Status #
# o *History ->(IU) *,StatusType,Status #
################################################################
#################################################################################
##EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF
## '''
## ##############################################################################
## # MISC FUNCTIONS
## ##############################################################################
## '''
## Check the booster ResourceStatusSystem.Utilities.ResourceStatusBooster
## def setMonitoredToBeChecked( self, monitoreds, granularity, name ):
## """
## Set LastCheckTime to 0 to monitored(s)
##
## :params:
## :attr:`monitoreds`: string, or a list of strings where each is a ValidElement:
## which granularity has to be set to be checked
##
## :attr:`granularity`: string, a ValidElement: from who this set comes
##
## :attr:`name`: string, name of Site or Resource
## """
##
## znever = datetime.min
##
## if type( monitoreds ) is not list:
## monitoreds = [ monitoreds ]
##
## for monitored in monitoreds:
##
## if monitored == 'Site':
##
## siteName = self.getGeneralName( granularity, name, monitored )[ 'Value' ]
## self.updateSiteStatus(siteName = siteName, lastCheckTime = znever )
##
## elif monitored == 'Service' :
##
## if granularity =='Site':
## serviceName = self.getMonitoredsList( 'Service', paramsList = [ 'ServiceName' ],
## siteName = name )[ 'Value' ]
## if type( serviceName ) is not list:
## serviceName = [ serviceName ]
## if serviceName != []:
### raise Exception, where( self, self.setMonitoredToBeChecked ) + " No services for site %s" %name
### else:
## serviceName = [ x[0] for x in serviceName ]
## self.updateServiceStatus( serviceName = serviceName, lastCheckTime = znever )
## else:
## serviceName = self.getGeneralName( granularity, name, monitored )[ 'Value' ]
## self.updateServiceStatus( serviceName = serviceName, lastCheckTime = znever )
##
## elif monitored == 'Resource':
##
## if granularity == 'Site' :
## resourceName = self.getMonitoredsList( 'Resource', paramsList = [ 'ResourceName' ],
## siteName = name )[ 'Value' ]
## if type( resourceName ) is not list:
## resourceName = [ resourceName ]
## if resourceName != []:
## #raise Exception, where( self, self.setMonitoredToBeChecked ) + " No resources for site %s" %name
## #else:
## resourceName = [ x[0] for x in resourceName ]
## self.updateResourceStatus( resourceName = resourceName, lastCheckTime = znever )
##
## elif granularity == 'Service' :
##
## #siteName = self.getGeneralName( granularity, name, 'Resource' )
## serviceType, siteName = name.split('@')
## gridSiteName = self.getGridSiteName('Site', siteName)[ 'Value' ]
##
## resourceName = self.getMonitoredsList( monitored, paramsList = [ 'ResourceName' ],
## gridSiteName = gridSiteName,
## serviceType = serviceType )[ 'Value' ]
## if type( resourceName ) is not list:
## resourceName = [ resourceName ]
## if resourceName != []:
## # raise Exception, where( self, self.setMonitoredToBeChecked ) + " No resources for service %s" %name
## # else:
## resourceName = [ x[0] for x in resourceName ]
## self.updateResourceStatus( resourceName = resourceName, lastCheckTime = znever )
##
## elif granularity == 'StorageElement':
## resourceName = self.getGeneralName( granularity, name, monitored )[ 'Value' ]
## self.updateResourceStatus( resourceName = resourceName, lastCheckTime = znever )
##
## # Put read and write together here... too much fomr copy/paste
## elif monitored == 'StorageElement':
##
## if granularity == 'Site':
##
## gridSiteName = self.getGridSiteName('Site', siteName)[ 'Value' ]
## SEName = self.getMonitoredsList( monitored, paramsList = [ 'StorageElementName' ],
## gridSiteName = gridSiteName )[ 'Value' ]
## if type( SEName ) is not list:
## SEName = [ SEName ]
## if SEName != []:
## #pass
## #else:
## SEName = [ x[0] for x in SEName ]
## self.updateStorageElementStatus( storageElementName = SEName, lastCheckTime = znever )
##
## elif granularity == 'Resource':
## SEName = self.getMonitoredsList( monitored, paramsList = [ 'StorageElementName' ],
## resourceName = name )[ 'Value' ]
## if type( SEName ) is not list:
## SEName = [ SEName ]
## if SEName == []:
## pass
### raise Exception, where(self, self.setMonitoredToBeChecked) + "No storage elements for resource %s" %name
## else:
## SEName = [ x[0] for x in SEName ]
## self.updateStorageElementStatus( storageElementName = SEName, lastCheckTime = znever )
##
## elif granularity == 'Service':
##
## serviceType, siteName = name.split('@')
## gridSiteName = self.getGridSiteName('Site', siteName)[ 'Value' ]
##
## SEName = self.getMonitoredsList( monitored, paramsList = [ 'StorageElementName' ],
## gridSiteName = gridSiteName )[ 'Value' ]#name.split('@').pop() )[ 'Value' ]
## if type( SEName ) is not list:
## SEName = [ SEName ]
## if SEName != []:
## #pass
### raise Exception, where(self, self.setMonitoredToBeChecked) + "No storage elements for service %s" %name
## #else:
## SEName = [ x[0] for x in SEName ]
## self.updateStorageElementStatus( storageElementName = SEName, lastCheckTime = znever )
################################################################################
#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF
class ResourceStatusDB( object ):
"""
The ResourceStatusDB class is a front-end to the ResourceStatusDB MySQL db.
It exposes four basic methods:
- insert
- update
- get
- delete
all them defined on the MySQL monkey class.
Moreover, there are a set of key-worded parameters that can be used, specially
on the getX and deleteX functions ( to know more, again, check the MySQL monkey
documentation ).
The DB schema has NO foreign keys, so there may be some small consistency checks,
called validators on the insert and update functions.
The simplest way to instantiate an object of type :class:`ResourceStatusDB`
is simply by calling
>>> rsDB = ResourceStatusDB()
This way, it will use the standard :mod:`DIRAC.Core.Base.DB`.
But there's the possibility to use other DB classes.
For example, we could pass custom DB instantiations to it,
provided the interface is the same exposed by :mod:`DIRAC.Core.Base.DB`.
>>> AnotherDB = AnotherDBClass()
>>> rsDB = ResourceStatusDB( DBin = AnotherDB )
Alternatively, for testing purposes, you could do:
>>> from mock import Mock
>>> mockDB = Mock()
>>> rsDB = ResourceStatusDB( DBin = mockDB )
Or, if you want to work with a local DB, providing it's mySQL:
>>> rsDB = ResourceStatusDB( DBin = [ 'UserName', 'Password' ] )
The ResourceStatusDB also exposes database Schema information, either on a
dictionary or on a MySQLSchema tree object.
- getSchema
- inspectSchema
Alternatively, we can access the MySQLSchema XML and tree as follows:
>>> rsDB = ResourceStatusDB()
>>> xml = rsDB.mm.SCHEMA
>>> tree = rsDB.mm.mSchema
>>> tree
>>> tree.GridSite
>>> tree.GridSite.GridTier
"""
def __init__( self, *args, **kwargs ):
"""Constructor."""
if len( args ) == 1:
if isinstance( args[ 0 ], str ):
maxQueueSize = 10
if isinstance( args[ 0 ], int ):
maxQueueSize = args[ 0 ]
elif len( args ) == 2:
maxQueueSize = args[ 1 ]
elif len( args ) == 0:
maxQueueSize = 10
if 'DBin' in kwargs.keys():
dbIn = kwargs[ 'DBin' ]
if isinstance( dbIn, list ):
from DIRAC.Core.Utilities.MySQL import MySQL
self.db = MySQL( 'localhost', dbIn[ 0 ], dbIn[ 1 ], 'ResourceStatusDB' )
else:
self.db = dbIn
else:
from DIRAC.Core.Base.DB import DB
self.db = DB( 'ResourceStatusDB', 'ResourceStatus/ResourceStatusDB', maxQueueSize )
self.mm = MySQLMonkey( self )
@CheckDBExecution
@ValidateDBTypes
def insert( self, params, meta ):
"""
Inserts args in the DB making use of kwargs where parameters such as
the table are specified ( filled automatically by the Client). In order to
do the insertion, it uses MySQLMonkey to do the parsing, execution and
error handling. Typically you will not pass kwargs to this function, unless
you know what are you doing and you have a very special use case.
:Parameters:
**params** - `dict`
arguments for the mysql query ( must match table columns ! ).
**meta** - `dict`
metadata for the mysql query. It must contain, at least, `table` key
with the proper table name.
:return: S_OK() || S_ERROR()
"""
return self.mm.insert( params, meta )
@CheckDBExecution
@ValidateDBTypes
def update( self, params, meta ):
"""
Updates row with values given on args. The row selection is done using the
default of MySQLMonkey ( column.primary or column.keyColumn ). It can be
modified using kwargs, but it is not explained here. The table keyword
argument is mandatory, and filled automatically by the Client. Typically
you will not pass kwargs to this function, unless you know what are you
doing and you have a very special use case.
:Parameters:
**params** - `dict`
arguments for the mysql query ( must match table columns ! ).
**meta** - `dict`
metadata for the mysql query. It must contain, at least, `table` key
with the proper table name.
:return: S_OK() || S_ERROR()
"""
return self.mm.update( params, meta )
@CheckDBExecution
@ValidateDBTypes
def get( self, params, meta ):
"""
Uses arguments to build conditional SQL statement ( WHERE ... ). If the
sql statement desired is more complex, you can use kwargs to interact with
the MySQLStatement parser and generate a more sophisticated query.
:Parameters:
**params** - `dict`
arguments for the mysql query ( must match table columns ! ).
**meta** - `dict`
metadata for the mysql query. It must contain, at least, `table` key
with the proper table name.
:return: S_OK() || S_ERROR()
"""
return self.mm.get( params, meta )
@CheckDBExecution
@ValidateDBTypes
def delete( self, params, meta ):
"""
Uses arguments to build conditional SQL statement ( WHERE ... ). If the
sql statement desired is more complex, you can use kwargs to interact with
the MySQLStatement parser and generate a more sophisticated query. There is
only one forbidden query, with all parameters None ( this would mean a query
of the type `DELETE * from TableName` ). The usage of kwargs is the same as in
the get function.
:Parameters:
**params** - `dict`
arguments for the mysql query ( must match table columns ! ).
**meta** - `dict`
metadata for the mysql query. It must contain, at least, `table` key
with the proper table name.
:return: S_OK() || S_ERROR()
"""
return self.mm.delete( params, meta )
@CheckDBExecution
def getSchema( self ):
"""
Returns a dictionary with database schema, this includes table and column
names. It has two variants, columns and keyUsage. The first one has at least,
as many keys as keyUsage, it is the complete schema. The second one is the
one used for the default updates and selects -- not taking into account
auto_increment fields, but taking into account primary and keyUsage fields.
:Parameters:
`None`
:return: S_OK()
"""
return { 'OK': True, 'Value' : self.mm.SCHEMA }
@CheckDBExecution
def inspectSchema( self ):
"""
Returns an object which represents the database schema and can be browsed.
>>> db = ResourceStatusDB()
>>> schema = db.inspectSchema()[ 'Value' ]
>>> schema
Schema 123:
<TableName1>,<TableName2>...
>>> schema.TableName1
Table TableName1:
<ColumnName1>,<ColumnName2>..
Every column has a few attributes ( primary, keyUsage, extra, position,
dataType and charMaxLen ).
:Parameters:
`None`
:return: S_OK()
"""
return { 'OK': True, 'Value' : self.mm.mSchema }
################################################################
# #
# VALIDATION ?? #
# #
# o Site ->(IU) SiteType #
# o Service ->(IU) ServiceType, Site #
# o Resource ->(IU) ResourceType, ServiceType #
# o StorageElement ->(IU) Resource #
# o GridSite ->(IU) __none__ #
# o *Status ->(IU) *,StatusType,Status #
# o *ScheduledStatus ->(IU) *,StatusType,Status #
# o *History ->(IU) *,StatusType,Status #
################################################################
#################################################################################
##EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF
## '''
## ##############################################################################
## # MISC FUNCTIONS
## ##############################################################################
## '''
## Check the booster ResourceStatusSystem.Utilities.ResourceStatusBooster
## def setMonitoredToBeChecked( self, monitoreds, granularity, name ):
## """
## Set LastCheckTime to 0 to monitored(s)
##
## :params:
## :attr:`monitoreds`: string, or a list of strings where each is a ValidElement:
## which granularity has to be set to be checked
##
## :attr:`granularity`: string, a ValidElement: from who this set comes
##
## :attr:`name`: string, name of Site or Resource
## """
##
## znever = datetime.min
##
## if type( monitoreds ) is not list:
## monitoreds = [ monitoreds ]
##
## for monitored in monitoreds:
##
## if monitored == 'Site':
##
## siteName = self.getGeneralName( granularity, name, monitored )[ 'Value' ]
## self.updateSiteStatus(siteName = siteName, lastCheckTime = znever )
##
## elif monitored == 'Service' :
##
## if granularity =='Site':
## serviceName = self.getMonitoredsList( 'Service', paramsList = [ 'ServiceName' ],
## siteName = name )[ 'Value' ]
## if type( serviceName ) is not list:
## serviceName = [ serviceName ]
## if serviceName != []:
### raise Exception, where( self, self.setMonitoredToBeChecked ) + " No services for site %s" %name
### else:
## serviceName = [ x[0] for x in serviceName ]
## self.updateServiceStatus( serviceName = serviceName, lastCheckTime = znever )
## else:
## serviceName = self.getGeneralName( granularity, name, monitored )[ 'Value' ]
## self.updateServiceStatus( serviceName = serviceName, lastCheckTime = znever )
##
## elif monitored == 'Resource':
##
## if granularity == 'Site' :
## resourceName = self.getMonitoredsList( 'Resource', paramsList = [ 'ResourceName' ],
## siteName = name )[ 'Value' ]
## if type( resourceName ) is not list:
## resourceName = [ resourceName ]
## if resourceName != []:
## #raise Exception, where( self, self.setMonitoredToBeChecked ) + " No resources for site %s" %name
## #else:
## resourceName = [ x[0] for x in resourceName ]
## self.updateResourceStatus( resourceName = resourceName, lastCheckTime = znever )
##
## elif granularity == 'Service' :
##
## #siteName = self.getGeneralName( granularity, name, 'Resource' )
## serviceType, siteName = name.split('@')
## gridSiteName = self.getGridSiteName('Site', siteName)[ 'Value' ]
##
## resourceName = self.getMonitoredsList( monitored, paramsList = [ 'ResourceName' ],
## gridSiteName = gridSiteName,
## serviceType = serviceType )[ 'Value' ]
## if type( resourceName ) is not list:
## resourceName = [ resourceName ]
## if resourceName != []:
## # raise Exception, where( self, self.setMonitoredToBeChecked ) + " No resources for service %s" %name
## # else:
## resourceName = [ x[0] for x in resourceName ]
## self.updateResourceStatus( resourceName = resourceName, lastCheckTime = znever )
##
## elif granularity == 'StorageElement':
## resourceName = self.getGeneralName( granularity, name, monitored )[ 'Value' ]
## self.updateResourceStatus( resourceName = resourceName, lastCheckTime = znever )
##
## # Put read and write together here... too much fomr copy/paste
## elif monitored == 'StorageElement':
##
## if granularity == 'Site':
##
## gridSiteName = self.getGridSiteName('Site', siteName)[ 'Value' ]
## SEName = self.getMonitoredsList( monitored, paramsList = [ 'StorageElementName' ],
## gridSiteName = gridSiteName )[ 'Value' ]
## if type( SEName ) is not list:
## SEName = [ SEName ]
## if SEName != []:
## #pass
## #else:
## SEName = [ x[0] for x in SEName ]
## self.updateStorageElementStatus( storageElementName = SEName, lastCheckTime = znever )
##
## elif granularity == 'Resource':
## SEName = self.getMonitoredsList( monitored, paramsList = [ 'StorageElementName' ],
## resourceName = name )[ 'Value' ]
## if type( SEName ) is not list:
## SEName = [ SEName ]
## if SEName == []:
## pass
### raise Exception, where(self, self.setMonitoredToBeChecked) + "No storage elements for resource %s" %name
## else:
## SEName = [ x[0] for x in SEName ]
## self.updateStorageElementStatus( storageElementName = SEName, lastCheckTime = znever )
##
## elif granularity == 'Service':
##
## serviceType, siteName = name.split('@')
## gridSiteName = self.getGridSiteName('Site', siteName)[ 'Value' ]
##
## SEName = self.getMonitoredsList( monitored, paramsList = [ 'StorageElementName' ],
## gridSiteName = gridSiteName )[ 'Value' ]#name.split('@').pop() )[ 'Value' ]
## if type( SEName ) is not list:
## SEName = [ SEName ]
## if SEName != []:
## #pass
### raise Exception, where(self, self.setMonitoredToBeChecked) + "No storage elements for service %s" %name
## #else:
## SEName = [ x[0] for x in SEName ]
## self.updateStorageElementStatus( storageElementName = SEName, lastCheckTime = znever )
################################################################################
#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF#EOF
class ResourceManagementDB(object):
"""
The ResourceManagementDB class is a front-end to the ResourceManagementDB MySQL db.
If exposes four basic methods:
- insert
- update
- get
- delete
all them defined on the MySQL monkey class.
Moreover, there are a set of key-worded parameters that can be used, specially
on the getX and deleteX functions ( to know more, again, check the MySQL monkey
documentation ).
The DB schema has NO foreign keys, so there may be some small consistency checks,
called validators on the insert and update functions.
The simplest way to instantiate an object of type :class:`ResourceManagementDB`
is simply by calling
>>> rmDB = ResourceManagementDB()
This way, it will use the standard :mod:`DIRAC.Core.Base.DB`.
But there's the possibility to use other DB classes.
For example, we could pass custom DB instantiations to it,
provided the interface is the same exposed by :mod:`DIRAC.Core.Base.DB`.
>>> AnotherDB = AnotherDBClass()
>>> rmDB = ResourceManagementDB( DBin = AnotherDB )
Alternatively, for testing purposes, you could do:
>>> from mock import Mock
>>> mockDB = Mock()
>>> rmDB = ResourceManagementDB( DBin = mockDB )
Or, if you want to work with a local DB, providing it's mySQL:
>>> rmDB = ResourceManagementDB( DBin = [ 'UserName', 'Password' ] )
The ResourceStatusDB also exposes database Schema information, either on a
dictionary or on a MySQLSchema tree object.
- getSchema
- inspectSchema
Alternatively, we can access the MySQLSchema XML and tree as follows:
>>> rmDB = ResourceManagementDB()
>>> xml = rmDB.mm.SCHEMA
>>> tree = rmDB.mm.mSchema
>>> tree
>>> tree.
>>> tree.PolicyResult.Name
"""
def __init__(self, *args, **kwargs):
"""Constructor."""
if len(args) == 1:
if isinstance(args[0], str):
maxQueueSize = 10
if isinstance(args[0], int):
maxQueueSize = args[0]
elif len(args) == 2:
maxQueueSize = args[1]
elif len(args) == 0:
maxQueueSize = 10
if "DBin" in kwargs.keys():
dbIn = kwargs["DBin"]
if isinstance(dbIn, list):
from DIRAC.Core.Utilities.MySQL import MySQL
self.db = MySQL("localhost", dbIn[0], dbIn[1], "ResourceManagementDB")
else:
self.db = dbIn
else:
from DIRAC.Core.Base.DB import DB
self.db = DB("ResourceManagementDB", "ResourceStatus/ResourceManagementDB", maxQueueSize)
self.mm = MySQLMonkey(self)
@CheckDBExecution
@ValidateDBTypes
def insert(self, params, meta):
"""
Inserts args in the DB making use of kwargs where parameters such as
the table are specified ( filled automatically by the Client). In order to
do the insertion, it uses MySQLMonkey to do the parsing, execution and
error handling. Typically you will not pass kwargs to this function, unless
you know what are you doing and you have a very special use case.
:Parameters:
**params** - `dict`
arguments for the mysql query ( must match table columns ! ).
**meta** - `dict`
metadata for the mysql query. It must contain, at least, `table` key
with the proper table name.
:return: S_OK() || S_ERROR()
"""
return self.mm.insert(params, meta)
@CheckDBExecution
@ValidateDBTypes
def update(self, params, meta):
"""
Updates row with values given on args. The row selection is done using the
default of MySQLMonkey ( column.primary or column.keyColumn ). It can be
modified using kwargs, but it is not explained here. The table keyword
argument is mandatory, and filled automatically by the Client. Typically
you will not pass kwargs to this function, unless you know what are you
doing and you have a very special use case.
:Parameters:
**params** - `dict`
arguments for the mysql query ( must match table columns ! ).
**meta** - `dict`
metadata for the mysql query. It must contain, at least, `table` key
with the proper table name.
:return: S_OK() || S_ERROR()
"""
return self.mm.update(params, meta)
@CheckDBExecution
@ValidateDBTypes
def get(self, params, meta):
"""
Uses arguments to build conditional SQL statement ( WHERE ... ). If the
sql statement desired is more complex, you can use kwargs to interact with
the MySQLStatement parser and generate a more sophisticated query.
:Parameters:
**params** - `dict`
arguments for the mysql query ( must match table columns ! ).
**meta** - `dict`
metadata for the mysql query. It must contain, at least, `table` key
with the proper table name.
:return: S_OK() || S_ERROR()
"""
return self.mm.get(params, meta)
@CheckDBExecution
@ValidateDBTypes
def delete(self, params, meta):
"""
Uses arguments to build conditional SQL statement ( WHERE ... ). If the
sql statement desired is more complex, you can use kwargs to interact with
the MySQLStatement parser and generate a more sophisticated query. There is
only one forbidden query, with all parameters None ( this would mean a query
of the type `DELETE * from TableName` ). The usage of kwargs is the same
as in the get function.
:Parameters:
**params** - `dict`
arguments for the mysql query ( must match table columns ! ).
**meta** - `dict`
metadata for the mysql query. It must contain, at least, `table` key
with the proper table name.
:return: S_OK() || S_ERROR()
"""
return self.mm.delete(params, meta)
@CheckDBExecution
def getSchema(self):
"""
Returns a dictionary with database schema, this includes table and column
names. It has two variants, columns and keyUsage. The first one has at least,
as many keys as keyUsage, it is the complete schema. The second one is the
one used for the default updates and selects -- not taking into account
auto_increment fields, but taking into account primary and keyUsage fields.
:Parameters:
`None`
:return: S_OK()
"""
return {"OK": True, "Value": self.mm.SCHEMA}
@CheckDBExecution
def inspectSchema(self):
"""
Returns an object which represents the database schema and can be browsed.
>>> db = ResourceManagementDB()
>>> schema = db.inspectSchema()[ 'Value' ]
>>> schema
Schema 123:
<TableName1>,<TableName2>...
>>> schema.TableName1
Table TableName1:
<ColumnName1>,<ColumnName2>..
Every column has a few attributes ( primary, keyUsage, extra, position,
dataType and charMaxLen ).
:Parameters:
`None`
:return: S_OK()
"""
return {"OK": True, "Value": self.mm.mSchema}