Score:0

What are the IOPS and physical block size for modern magnetic tapes?

jp flag

It's obvious that magnetic tape media performs IO operations as well as HDDs and SSDs, although this storage device is not designed for a random access pattern.

LTO standards define maximal data transfer rate (also known as throughput) measured in bytes/sec for every LTO generation. But I wasn't able to find any information about IO rate with exception to this paper claiming that

Tape drive is available with 1000 IOPS

Also I'm curious if some kind of elementary entity similar to HDD's sector and SSD's block of pages exists in a magnetic tapes world, because its size directly impacts throughput:

  Throughput = IOPS * BlockSize
Score:3
ru flag

Physical tape is a sequential medium, so forget about (random access) IOPS. I've got some 90 seconds maximum rewind time in my head for LTO. Translated to 45 seconds mean access time, you'd get .02 IOPS.

It's the sustained transfer rate that's of interest. LTO-9 runs at up to 400 MB/s.

Perhaps the authors are using some exotic definition for IOPS - usually the term is used for random access. If you use it to count sequential blocks, you'd get around 6000 blocks/s with average sized 64 KiB blocks for LTO-9 (block size isn't fixed for most tape drives).

Virtual tapes or libraries are a completely different story. They're not hampered by tape's physical limitations, so you could get several hundred or thousand IOPS when backed by HDDs, or even millions when backed by SSDs.

I sit in a Tesla and translated this thread with Ai:

mangohost

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.