* Add more FAQs

* Update

* Update

* Update

* Update

* Add "Fault Diagnosis and Analysis" doc

* Update

* Update

* Update

Redis client will not retry the request after read timeout, so we increase
read timeout from 3 seconds to 30 seconds. The default of write timeout is
read timeout, so we also change that to 5 seconds (will be retried).

* Add cache management doc

* Update

* Update

* Update

* Update K8s doc link in README_CN.md

* Update

* WIP: S3 gateway

* upgrade minio

* build gateway

* use CopyFileRange to speedup Copy and MultipartUpload

* add subcommand

* fix errno handling

* MinIO require Go 1.14+

* fix lint

* rollback change in travis

* fix lint

* add how to use on kubernetes doc

* add doc for installation with kubectl

* Add end-to-end checksum for S3/OSS/COS

* add license header

* fix lint

* fix lint

* refresh OSS token

* fix ci

* refresh token inmediately

* fix lint

* Update data encryption related docs

* Update

* encrypt data at-rest

* add docs

* fix docs

* fix lint

* Export metrics to Prometheus

* ignore error

* listen in background

* add metrics for request to object store

* metrics for open files/dirs

* fix lint

* Add doc: Comparison with Others

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update comparison_with_others.md

change the order
Co-authored-by: NDavies Liu <davies@juicedata.io>

* Update doc: add recommended managed Redis service

* Update

* Update

* Fix typo

* build static binary

* use musl-libc to build static linked binary for linux amd64 platform

* fix

* fix
Co-authored-by: Nliyong <chnliyong@gmail.com>

* support concat in Java SDK

* fix lint

* remove logging

* support copy_file_range

* cleanup slices with negative refs

* Add doc: Redis best practices

* Update TOC

* Update TOC

* Add doc: mount JuiceFS at boot

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Update

* Restore pkg/object/README.md