diff --git a/infra/docs/02-kata-runtime.md b/infra/docs/02-kata-runtime.md
index 069d600f4..33e3137f3 100644
--- a/infra/docs/02-kata-runtime.md
+++ b/infra/docs/02-kata-runtime.md
@@ -264,7 +264,7 @@ image  = "/opt/kata/share/kata-containers/kata-containers.img"
 machine_type = "q35"
 rootfs_type  = "ext4"
 cpu_features  = "pmu=off"
-kernel_params = "cgroup_no_v1=all systemd.unified_cgroup_hierarchy=1"
+kernel_params = "cgroup_no_v1=all systemd.unified_cgroup_hierarchy=1 sysctl.fs.nr_open=8388608"
 
 default_vcpus    = 2
 default_maxvcpus = 0
@@ -275,7 +275,7 @@ memory_slots     = 10
 shared_fs        = "virtio-fs"
 virtio_fs_daemon = "/opt/kata/libexec/virtiofsd"
 virtio_fs_cache  = "auto"
-virtio_fs_extra_args = ["--thread-pool-size=4", "--announce-submounts"]
+virtio_fs_extra_args = ["--thread-pool-size=4", "--announce-submounts", "--rlimit-nofile=8388608"]
 
 disable_block_device_use = true
 block_device_driver = "virtio-scsi"
@@ -300,9 +300,21 @@ sandbox_cgroup_only = false
 `kata-agent` baked in as PID 1's manager). These three are the entire guest —
 none of them is the host kernel, which is the whole point. `machine_type =
 "q35"` is the modern PCIe QEMU machine (needed for PCIe hotplug);
-`cpu_features = "pmu=off"` disables the virtual perf-monitoring unit (avoids
-spurious PMU passthrough issues). `kernel_params` forces cgroup v2-only in the
-guest, matching a modern systemd userspace.
+`cpu_features = "pmu=off"` disables the virtual perf-monitoring unit and avoids
+spurious PMU passthrough issues. The first two `kernel_params` entries force
+cgroup v2-only in the guest, matching a modern systemd userspace.
+
+### Open-file ceilings — Bazel sandbox headroom
+
+Two limits protect each runner. `sysctl.fs.nr_open=8388608` raises the guest
+open-file ceiling from 1,048,576. New runner containers inherit this value for
+both soft and hard open-file resource limits (`RLIMIT_NOFILE`).
+
+`--rlimit-nofile=8388608` raises the host `virtiofsd` process ceiling.
+Virtiofsd otherwise caps itself at 1,000,000 descriptors. Cold Bazel builds
+index large Zig and xwin trees through this daemon. The daemon exhausted its old
+ceiling and returned `EMFILE` to Bazel. Linux grows descriptor tables on demand,
+so unused headroom has no fixed allocation.
 
 ### vCPU / memory sizing — hotplug from pod requests/limits
 
diff --git a/infra/tune-kata-runtime.sh b/infra/tune-kata-runtime.sh
index 8c782a523..2766d7c24 100755
--- a/infra/tune-kata-runtime.sh
+++ b/infra/tune-kata-runtime.sh
@@ -1,8 +1,8 @@
 #!/usr/bin/env bash
-# Patch the live Kata QEMU config on the CI host to match the runner pod's
-# guaranteed boot shape and a larger virtiofsd worker pool, then smoke-test that
-# a new kata-qemu pod still boots. Driven over SSH from this repo so the desired
-# values stay version-controlled.
+# Patch the live Kata QEMU config to match the runner pod's guaranteed boot
+# shape, raise guest and host open-file limits, and enlarge the virtiofsd worker pool.
+# The script runs over SSH, which keeps the desired values version-controlled.
+# It then smoke-tests a new kata-qemu pod.
 #
 # Usage:
 #   CI_HOST=my-ci-host ./infra/tune-kata-runtime.sh
@@ -16,6 +16,7 @@
 #   BOOT_VCPUS             Kata default_vcpus                            [2]
 #   BOOT_MEMORY_MIB        Kata default_memory (MiB)                     [4096]
 #   VIRTIOFSD_THREAD_POOL  virtiofsd --thread-pool-size                  [4]
+#   OPEN_FILE_LIMIT        guest and virtiofsd open-file limit            [8388608]
 set -euo pipefail
 
 : "${CI_HOST:?set CI_HOST to the ssh target of your CI host, e.g. CI_HOST=my-ci-host}"
@@ -26,10 +27,11 @@ ARC_NAMESPACE="${ARC_NAMESPACE:-arc-runners}"
 BOOT_VCPUS="${BOOT_VCPUS:-2}"
 BOOT_MEMORY_MIB="${BOOT_MEMORY_MIB:-4096}"
 VIRTIOFSD_THREAD_POOL="${VIRTIOFSD_THREAD_POOL:-4}"
+OPEN_FILE_LIMIT="${OPEN_FILE_LIMIT:-8388608}"
 
 ssh "$CI_HOST" bash -s -- \
   "$KATA_CONFIG_REMOTE" "$KUBECONFIG_REMOTE" "$ARC_RELEASE" "$ARC_NAMESPACE" \
-  "$BOOT_VCPUS" "$BOOT_MEMORY_MIB" "$VIRTIOFSD_THREAD_POOL" <<'REMOTE'
+  "$BOOT_VCPUS" "$BOOT_MEMORY_MIB" "$VIRTIOFSD_THREAD_POOL" "$OPEN_FILE_LIMIT" <<'REMOTE'
 set -euo pipefail
 KATA_CONFIG="$1"
 export KUBECONFIG="$2"
@@ -38,12 +40,13 @@ ARC_NAMESPACE="$4"
 BOOT_VCPUS="$5"
 BOOT_MEMORY_MIB="$6"
 THREAD_POOL="$7"
+OPEN_FILE_LIMIT="$8"
 
 backup="${KATA_CONFIG}.bak.$(date +%Y%m%d-%H%M%S)"
 cp "$KATA_CONFIG" "$backup"
 echo "==> backup: $backup"
 
-python3 - "$KATA_CONFIG" "$BOOT_VCPUS" "$BOOT_MEMORY_MIB" "$THREAD_POOL" <<'PY'
+python3 - "$KATA_CONFIG" "$BOOT_VCPUS" "$BOOT_MEMORY_MIB" "$THREAD_POOL" "$OPEN_FILE_LIMIT" <<'PY'
 from pathlib import Path
 import re
 import sys
@@ -51,21 +54,43 @@ path = Path(sys.argv[1])
 boot_vcpus = sys.argv[2]
 boot_mem = sys.argv[3]
 thread_pool = sys.argv[4]
+fd_limit_value = int(sys.argv[5])
+if fd_limit_value < 1:
+    raise SystemExit("OPEN_FILE_LIMIT must be positive")
+fd_limit = str(fd_limit_value)
 text = path.read_text()
 replacements = [
     (r'(^\s*default_vcpus\s*=\s*)\d+', rf'\g<1>{boot_vcpus}'),
     (r'(^\s*default_memory\s*=\s*)\d+', rf'\g<1>{boot_mem}'),
-    (r'(^\s*virtio_fs_extra_args\s*=\s*)\[[^\]]*\]', rf'\g<1>["--thread-pool-size={thread_pool}", "--announce-submounts"]'),
+    (r'(^\s*virtio_fs_extra_args\s*=\s*)\[[^\]]*\]', rf'\g<1>["--thread-pool-size={thread_pool}", "--announce-submounts", "--rlimit-nofile={fd_limit}"]'),
 ]
 for pattern, replacement in replacements:
     text, n = re.subn(pattern, replacement, text, count=1, flags=re.MULTILINE)
     if n != 1:
         raise SystemExit(f"failed to patch {pattern}")
+kernel_pattern = r'^(\s*kernel_params\s*=\s*")([^"]*)(".*)$'
+def update_kernel_params(match):
+    params = [
+        param
+        for param in match.group(2).split()
+        if not param.startswith("sysctl.fs.nr_open=")
+    ]
+    params.append(f"sysctl.fs.nr_open={fd_limit}")
+    return f'{match.group(1)}{" ".join(params)}{match.group(3)}'
+text, n = re.subn(
+    kernel_pattern,
+    update_kernel_params,
+    text,
+    count=1,
+    flags=re.MULTILINE,
+)
+if n != 1:
+    raise SystemExit("failed to patch kernel_params")
 path.write_text(text)
 PY
 
 echo "==> active Kata knobs"
-grep -nE 'default_vcpus|default_memory|virtio_fs_extra_args' "$KATA_CONFIG"
+grep -nE 'kernel_params|default_vcpus|default_memory|virtio_fs_extra_args' "$KATA_CONFIG"
 
 image="$(kubectl get autoscalingrunnerset "$ARC_RELEASE" -n "$ARC_NAMESPACE" -o jsonpath='{.spec.template.spec.containers[0].image}')"
 pod="kata-runtime-smoke-$(date +%H%M%S)"
@@ -77,5 +102,26 @@ kubectl run "$pod" -n "$ARC_NAMESPACE" --restart=Never --image="$image" \
   --command -- bash -lc 'sleep 120' >/dev/null
 kubectl wait --for=condition=Ready "pod/$pod" -n "$ARC_NAMESPACE" --timeout=120s >/dev/null
 kubectl exec -n "$ARC_NAMESPACE" "$pod" -- bash -lc 'bun --version; rustc --version | head -1'
+fd_probe="$(kubectl exec -n "$ARC_NAMESPACE" "$pod" -- bash -lc \
+  'printf "%s %s %s" "$(cat /proc/sys/fs/nr_open)" "$(ulimit -Sn)" "$(ulimit -Hn)"')"
+read -r guest_nr_open soft_limit hard_limit <<<"$fd_probe"
+if [[ "$guest_nr_open" != "$OPEN_FILE_LIMIT" ||
+      "$soft_limit" != "$OPEN_FILE_LIMIT" ||
+      "$hard_limit" != "$OPEN_FILE_LIMIT" ]]; then
+  echo "fd limit mismatch: nr_open=$guest_nr_open soft=$soft_limit hard=$hard_limit expected=$OPEN_FILE_LIMIT" >&2
+  exit 1
+fi
+echo "guest fd limits: nr_open=$guest_nr_open soft=$soft_limit hard=$hard_limit"
+sandbox_id="$(k3s crictl pods --name "$pod" -q)"
+virtiofsd_pid="$(pgrep -fo "/opt/kata/libexec/virtiofsd.*sandboxes/${sandbox_id}/")"
+read -r virtiofsd_soft_limit virtiofsd_hard_limit < <(
+  awk '/^Max open files/ { print $4, $5 }' "/proc/$virtiofsd_pid/limits"
+)
+if [[ "$virtiofsd_soft_limit" != "$OPEN_FILE_LIMIT" ||
+      "$virtiofsd_hard_limit" != "$OPEN_FILE_LIMIT" ]]; then
+  echo "virtiofsd fd limit mismatch: soft=$virtiofsd_soft_limit hard=$virtiofsd_hard_limit expected=$OPEN_FILE_LIMIT" >&2
+  exit 1
+fi
+echo "virtiofsd fd limits: soft=$virtiofsd_soft_limit hard=$virtiofsd_hard_limit"
 echo "OK: kata-qemu still boots after tuning"
 REMOTE