I also understood your use case, and it's valid :-). There may be a corner case for somethings. Like pool import.
One way to deal with the pool import, is to manually run them sequentially. It will take longer, but it's more controlled. Another tuning step you might try, is to flush the RAM cache after each import. Next, since these are in essence R/O files, I'd disable "atime" in the dataset level. Did not see a reference to that feature in the E-Mail thread. Last, I'd look at the "primarycache" options. If your use case has the files read / streamed occasionally, then perhaps using "primarycache=metadata" might make sense. Especially if the files read are flushed out of RAM cache / ARC so quickly that it makes no sense to put them there in the first place. -- Lady Galadriel ------------------------------------------ openzfs: openzfs-developer Permalink: https://openzfs.topicbox.com/groups/developer/T10533b84f9e1cfc5-M011b45c3613c612209fc6cf5 Delivery options: https://openzfs.topicbox.com/groups/developer/subscription
