Static file reaches a certain size, quantity and capacity under the condition of high concurrency, sometimes leads to web application load time is too long, or directly for a long time no response, want to know what kind of way of thinking can be targeted to optimize the retrieval of the mass small file retrieval algorithm, the thought of a small file into a large file, then big small files in the file number (similar to a disk partition track), should be a waste of storage space, and divide different needs according to the actual size of the file interval, don't know whether it is feasible to, bosses, some directions or provide some other ideas,

CodePudding user response:

Try Mangodb and so on

CodePudding user response:

reference 1st floor assky124 response:

try Mangodb like

Well bosses said should be a certain degree of speed, different database retrieval algorithm affects the retrieval speed, but scale to a certain degree should be a drop in the bucket, the file itself to begin, don't know bosses think of my idea is feasible

CodePudding user response:

If conditions permit, to construct multiple mirror server
You can refer to the aws s3 and cloudfront architecture