用 get_or_create 多线程创建数据时会产生重复数据;
方法的源码如下:
def get_or_create(self, defaults=None, **kwargs):
"""
Looks up an object with the given kwargs, creating one if necessary.
Returns a tuple of (object, created), where created is a boolean
specifying whether an object was created.
"""
lookup, params = self._extract_model_params(defaults, **kwargs)
# The get() needs to be targeted at the write database in order
# to avoid potential transaction consistency problems.
self._for_write = True
try:
return self.get(**lookup), False
except self.model.DoesNotExist:
return self._create_object_from_params(lookup, params)
问题解决:
在创建表的时候指定联合索引;
配合事务一起使用;
在model层加限制unique或 unique_together
例子如下:
class PangleData(models.Model):
time_zone = models.CharField(max_length=3, verbose_name='Timezone')
currency = models.CharField(max_length=32, verbose_name='Currency')
region = models.CharField(max_length=6, verbose_name='Region')
app_id = models.IntegerField(verbose_name='App id on Pangle platform')
app_name = models.CharField(max_length=32, verbose_name='App Name on Pangle platform')
ad_slot_id = models.BigIntegerField(verbose_name='Placement id on Pangle platform')
ad_slot_type = models.IntegerField(verbose_name='Also known as Ad type')
package_name = models.CharField(max_length=32, verbose_name='Package name')
request = models.IntegerField(verbose_name='request')
P_return = models.IntegerField(verbose_name='return')
fill_rate = models.FloatField(null=True, verbose_name='fill_rate')
show = models.IntegerField(verbose_name='Impressions')
click = models.IntegerField(verbose_name='click')
click_rate = models.FloatField(verbose_name='Click Through Rate')
revenue = models.FloatField(verbose_name='Estimated Revenue')
ecpm = models.FloatField(verbose_name='Estimated Ecpm')
media_name = models.CharField(max_length=12, verbose_name='Account name')
code_name = models.CharField(max_length=64, verbose_name='Slot name on Pangle platform')
os = models.CharField(max_length=12, verbose_name='Operating system')
use_mediation = models.IntegerField(verbose_name='Is use pangle mediation. 0 (Not) 、1 (Yes)')
bidding_type = models.IntegerField(verbose_name='RIT bidding type. 0 (standard) 、1 (in-app bidding) 、'
'2 (client bidding)')
date = models.DateField(null=True, verbose_name='Date')
media_m_ssr = models.IntegerField(verbose_name='media_m_ssr')
class Meta:
unique_together = ['region', 'app_id', 'ad_slot_id']