Frequently asked questions about ONTAP FlexCache write-back
This FAQ can help if you are looking for a quick answer to a question.
Both the cache and the origin must be running ONTAP 9.15.1 or later. It is strongly recommended that you run the latest P release. Engineering is constantly improving the performance and functionality of write-back-enabled caches.
Yes. The origin has equal right to the data as any of the caches. If an operation is executed on a file that requires the file to be evicted from the cache, or a lock/data delegation to be revoked, the client at the cache might see a delay accessing the file.
Yes. Every cache and the origin can have independent QoS policies applied. This will have no direct effect on any write-back initiated intercluster traffic. Indirectly, you can slow down intercluster write-back traffic by QoS limiting the front-end traffic at the write-back-enabled cache.
Yes. Multi-protocol is fully supported at write-back-enabled FlexCaches. Currently, NFSv4.2 and S3 are not supported by FlexCache operating in write-around or write-back mode.
SMB alternate data streams (ADS) are supported, but not accelerated by write-back. The write to the ADS is forwarded to the origin, incurring the penalty of the WAN latency. The write also evicts the main file the ADS is a part of from the cache.
Yes. All you have to do is toggle the is-writeback-enabled
flag in the flexcache modify
command.