From 9ec9b89bf73985e509ef358325098844c2e1af43 Mon Sep 17 00:00:00 2001
From: Liang Yan <lyan@coreweave.com>
Date: Mon, 17 Aug 2026 12:27:54 -0400
Subject: [PATCH] runtime: batch ps -p in list -f instead of one call per entry

runtime::list() calls "ps -p" separately for each rootfs entry, in
alphabetical order, after the initial lsns scan that found the pids.
Entries later in that order get checked further away in time from
when lsns saw them running, so the staleness window #126/#220 already
deal with is not fixed size, it grows with the number of running
containers.

Measured on an idle VM with N containers running: the gap between the
lsns scan and each entry per-entry ps call went from ~46ms (1st entry)
to ~286ms (40th entry), growing roughly linearly with N. Batching a
single "ps -p" call right after the lsns scan and having the rest of
the function work off that one snapshot keeps this at a flat ~40ms
regardless of N.

Ran the #126 repro (owner exits mid-list) 8x against this, still
degrades gracefully to a name-only row like #220 intended, no
regression there.

Signed-off-by: Liang Yan <lyan@coreweave.com>
---
 src/runtime.sh | 41 ++++++++++++++++++++++++++++++-----------
 1 file changed, 30 insertions(+), 11 deletions(-)

diff --git a/src/runtime.sh b/src/runtime.sh
index 2d04567..40d5134 100644
--- a/src/runtime.sh
+++ b/src/runtime.sh
@@ -584,23 +584,42 @@ runtime::list() {
         fi
     done
 
+    # Snapshot ps/comm for every pid right after the lsns scan, in one shot,
+    # instead of one "ps -p" per entry below (see #126: that per-entry call
+    # is a wider version of the same race, and it gets wider with more
+    # containers since entries are checked one by one in sorted order).
+    local all_pids=() pid= comm= rest=
+    declare -A pid_info=()
+    for name in "${!info[@]}"; do
+        all_pids+=(${info["${name}"]})
+    done
+    if [ "${#all_pids[@]}" -gt 0 ]; then
+        while IFS=$'\t' read -r pid rest; do
+            comm=
+            read -r comm < "/proc/${pid}/comm" 2> /dev/null
+            pid_info["${pid}"]="${comm}"$'\t'"${rest}"
+        done < <(ps -p "${all_pids[*]}" --no-headers -o pid:1,stat:1,stime:1,etime:1,mntns:1,userns:1,command:1 2> /dev/null \
+                   | awk '{ printf "%s\t%s\t%s\t%s\t%s\t%s\t%s\n", $1, $2, $3, $4, $5, $6, substr($0, index($0, $7)) }')
+    fi
+
     # List all the rootfs entries and their respective processes.
     {
         printf "NAME\tPID\tCOMM\tSTATE\tSTARTED\tTIME\tMNTNS\tUSERNS\tCOMMAND\n"
         for name in $(printf "%s\n" "${!info[@]}" | sort); do
             entry=(${info["${name}"]})
-            if [ "${#entry[@]}" -eq 0 ]; then
-                printf "%s\n" "${name}"
-            else
-                ps -p "${entry[*]}" --no-headers -o pid:1,stat:1,stime:1,etime:1,mntns:1,userns:1,command:1 \
-                  | awk -v name="${name}" '{
-                      getline comm < ("/proc/"$1"/comm")
-                      printf (NR==1) ? "%s\t" : " \t", name
-                      printf "%s\t%s\t%s\t%s\t%s\t%s\t%s\t", $1, comm, $2, $3, $4, $5, $6
-                      print substr($0, index($0, $7))
-                      close("/proc/"$1"/comm")
-                  }' 2>/dev/null || printf "%s\n" "${name}"
+            local first=1
+            if [ "${#entry[@]}" -gt 0 ]; then
+                for pid in "${entry[@]}"; do
+                    [ -n "${pid_info["${pid}"]+x}" ] || continue
+                    if [ "${first}" -eq 1 ]; then
+                        printf "%s\t%s\t%s\n" "${name}" "${pid}" "${pid_info["${pid}"]}"
+                        first=0
+                    else
+                        printf " \t%s\t%s\n" "${pid}" "${pid_info["${pid}"]}"
+                    fi
+                done
             fi
+            [ "${first}" -eq 1 ] && printf "%s\n" "${name}"
         done
     } | column -t -s $'\t'
 }