Scenario is this: bulk update5000 data, have been running (per batch of 5000 continuously updated data)
Found that segment produce dozens of hundreds of 10 KB per second (assuming I 10 MB per batch data, should not produce a few 2 MB segment?),
Each batch bulk update dozens of seconds, even hundreds of seconds, too slow, I am not high concurrency of the single thread to run
The following figure
Environmental 8-core CPU 32 gb of memory hard disk 300 g, three machines,
160 million document, a total of 60 g harddisk
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.