nfsiostat(8)

💅 nfsiostat — the manual page, but make it slay

okay so bestie, you're looking at nfsiostat, a page from the System Manager's Manual. It's giving "iostat but for NFS" and honestly? iconic. Let's get into it 💖

---

✨ NAME ✨

nfsiostat — emulate iostat for NFS mount points using /proc/self/mountstats

---

🌸 SYNOPSIS 🌸

nfsiostat [[<interval>] [<count>]] [<options>][<mount_point>]

---

🦋 DESCRIPTION 🦋

The nfsiostat command shows you NFS client statistics, per mount. That's the whole vibe. 🫶

• <interval> — this is how many seconds you want between each report, fr fr. The first report is special though: it holds stats from the time each file system got mounted. Every report after that? Just the stats collected during the interval since the previous report. So it's like, a fresh lil slice each time. ✨
• <count> — if you pass <count>, that value decides how many reports you get, spaced out at <interval> seconds apart. And if you give an interval but no <count>? Then it just keeps generating reports continuously forever, bestie. No cap. 😭
• <options> — defined below 👇 (they're the whole moment)
• <mount_point> — name one or more <mount point>s and you'll only see stats for those. Name none? Then every NFS mount point on the client gets listed. You get the whole collection. 🎀

---

🧚 what each column actually means (iykyk)

• op/s — the number of operations per second.
• rpc bklog — the length of the backlog queue.
• kB/s (MB/s) — the number of kB (or MB) written/read per second.
• kB/op (MB/op) — the number of kB (or MB) written/read per each operation.
• retrans — the number of retransmissions.
• avg RTT (ms) — the duration from the time the client's kernel sends the RPC request until the time it receives the reply. That's the round trip, babe. ✈️
• avg exe (ms) — the duration from the time the NFS client does the RPC request to its kernel until the RPC request is completed. Heads up: this includes the RTT time from above. So it's RTT plus extra. 👑
• avg queue (ms) — the duration from the time the NFS client created the RPC request task to the time the request is transmitted. The lil waiting room era. 💅
• errors — the number of operations that completed with an error status (status < 0). Important tea: this count is only available on kernels with RPC iostats version 1.1 or above. So if you don't see it, it's not you, it's your kernel. 🥲

also real quick: if you use an interval as an argument to nfsiostat, then the difference from the previous interval gets displayed. If you don't? The results are from the time the share was mounted. Big difference, bestie, don't get caught slippin'. ⭐

---

🔥 OPTIONS 🔥

• -a or --attr — displays statistics related to the attribute cache ✨
• -d or --dir — displays statistics related to directory operations 📁
• -h or --help — shows help message and exit 🫶
• -l LIST or --list=LIST — only print stats for the first LIST mount points
• -m or --megabytes — display throughput in megabytes per second 💅
• -p or --page — displays statistics related to the page cache 📄
• -s or --sort — sort NFS mount points by ops/second 🔥
• --version — show program's version number and exit 🎀

---

🍓 FILES 🍓

/proc/self/mountstats

---

🌈 SEE ALSO 🌈

iostat(8), mountstats(8), nfsstat(8)

---

👑 AUTHOR 👑

Chuck Lever <chuck.lever@oracle.com>

---

and that's the whole page, bestie!! written by Chuck Lever, dated 15 Apr 2010, and it's the nfsiostat(8) page from the System Manager's Manual. slay, go check your mountstats, ily 💖🦋✨