summaryrefslogtreecommitdiff
path: root/src
diff options
context:
space:
mode:
authorPaul Buetow <paul@buetow.org>2026-06-14 00:00:10 +0300
committerPaul Buetow <paul@buetow.org>2026-06-14 00:00:10 +0300
commitf9291765b363809ff4e75ff441d374ecef3ad40a (patch)
tree9639677a5908469e16e67ccf0641a12d3f703c6e /src
parentd4971998edae804c39797486be7a342d43e61e5c (diff)
km0 harden parallel job pool
Diffstat (limited to 'src')
-rw-r--r--src/lib/album.source.sh60
-rw-r--r--src/lib/image.source.sh21
-rw-r--r--src/lib/job-pool.source.sh130
3 files changed, 178 insertions, 33 deletions
diff --git a/src/lib/album.source.sh b/src/lib/album.source.sh
index beec5fd..2036591 100644
--- a/src/lib/album.source.sh
+++ b/src/lib/album.source.sh
@@ -394,19 +394,36 @@ create_all_photo_derivatives() {
local -r blurs_dir="$1"; shift
local -i failed=0
local -a image_job_pids=()
+ # Passed by name to wait_for_image_job_slot and wait_for_image_jobs.
+ # shellcheck disable=SC2034
+ local -A image_job_labels=()
+ # Passed by name to wait_for_image_job_slot and wait_for_image_jobs.
+ # shellcheck disable=SC2034
+ local -A image_job_statuses=()
local photo
while IFS= read -r photo; do
- wait_for_image_job_slot image_job_pids failed
+ wait_for_image_job_slot \
+ image_job_pids \
+ image_job_statuses \
+ image_job_labels \
+ failed
create_photo_derivatives "$photos_dir" "$thumbs_dir" "$blurs_dir" \
"$photo" &
image_job_pids+=("$!")
+ # Read through a nameref in the job-pool helpers.
+ # shellcheck disable=SC2034
+ image_job_labels["$!"]="image derivative job for photo $photo"
done < <(
find "$DIST_DIR/$photos_dir" -maxdepth 1 -type f -printf '%f\n' \
| sort
)
- wait_for_image_jobs image_job_pids failed
+ wait_for_image_jobs \
+ image_job_pids \
+ image_job_statuses \
+ image_job_labels \
+ failed
if (( failed != 0 )); then
return 1
fi
@@ -633,10 +650,20 @@ queue_album_view_render_job() {
local -r photo="$1"; shift
# shellcheck disable=SC2178
local -n render_job_pids_ref="$1"; shift
+ # Passed by name to wait_for_template_render_job_slot.
+ # shellcheck disable=SC2034,SC2178
+ local -n render_job_statuses_ref="$1"; shift
+ # Passed by name to wait_for_template_render_job_slot and assigned below.
+ # shellcheck disable=SC2034,SC2178
+ local -n render_job_labels_ref="$1"; shift
# shellcheck disable=SC2178
local -n render_failed_ref="$1"; shift
- wait_for_template_render_job_slot render_job_pids_ref render_failed_ref
+ wait_for_template_render_job_slot \
+ render_job_pids_ref \
+ render_job_statuses_ref \
+ render_job_labels_ref \
+ render_failed_ref
render_photo_view_and_details \
"$photos_dir" \
"$blurs_dir" \
@@ -647,15 +674,26 @@ queue_album_view_render_job() {
"$preview_num" \
"$photo" &
render_job_pids_ref+=("$!")
+ render_job_labels_ref["$!"]="template render job for photo $photo"
}
wait_for_album_view_render_jobs() {
# shellcheck disable=SC2178
local -n render_job_pids_ref="$1"; shift
+ # Passed by name to wait_for_template_render_jobs.
+ # shellcheck disable=SC2034,SC2178
+ local -n render_job_statuses_ref="$1"; shift
+ # Passed by name to wait_for_template_render_jobs.
+ # shellcheck disable=SC2034,SC2178
+ local -n render_job_labels_ref="$1"; shift
# shellcheck disable=SC2178
local -n render_failed_ref="$1"; shift
- wait_for_template_render_jobs render_job_pids_ref render_failed_ref
+ wait_for_template_render_jobs \
+ render_job_pids_ref \
+ render_job_statuses_ref \
+ render_job_labels_ref \
+ render_failed_ref
if (( render_failed_ref != 0 )); then
return 1
fi
@@ -682,6 +720,12 @@ render_album_pages() {
# Passed by name to queue_album_view_render_job.
# shellcheck disable=SC2034
local -a render_job_pids=()
+ # Passed by name to queue_album_view_render_job.
+ # shellcheck disable=SC2034
+ local -A render_job_labels=()
+ # Passed by name to queue_album_view_render_job.
+ # shellcheck disable=SC2034
+ local -A render_job_statuses=()
# Passed by name to record_rendered_view_page and render_view_redirects.
# shellcheck disable=SC2034
local -A rendered_last_views=()
@@ -728,13 +772,19 @@ render_album_pages() {
"$i" \
"$photo" \
render_job_pids \
+ render_job_statuses \
+ render_job_labels \
render_failed
record_rendered_view_page rendered_view_pages rendered_last_views \
"$num" "$i"
done < <(album_photo_files "$photos_dir")
finish_preview_page "$name" "$html_dir" "$backhref" "$tarball_name"
- if ! wait_for_album_view_render_jobs render_job_pids render_failed; then
+ if ! wait_for_album_view_render_jobs \
+ render_job_pids \
+ render_job_statuses \
+ render_job_labels \
+ render_failed; then
return 1
fi
render_view_redirects "$html_dir" rendered_view_pages rendered_last_views
diff --git a/src/lib/image.source.sh b/src/lib/image.source.sh
index c190b87..f7cb715 100644
--- a/src/lib/image.source.sh
+++ b/src/lib/image.source.sh
@@ -66,15 +66,32 @@ warn_unsupported_incoming_files() {
scalephotos() {
local -i failed=0
local -a image_job_pids=()
+ # Passed by name to wait_for_image_job_slot and wait_for_image_jobs.
+ # shellcheck disable=SC2034
+ local -A image_job_labels=()
+ # Passed by name to wait_for_image_job_slot and wait_for_image_jobs.
+ # shellcheck disable=SC2034
+ local -A image_job_statuses=()
local photo
while IFS= read -r photo; do
- wait_for_image_job_slot image_job_pids failed
+ wait_for_image_job_slot \
+ image_job_pids \
+ image_job_statuses \
+ image_job_labels \
+ failed
scale_photo "$photo" &
image_job_pids+=("$!")
+ # Read through a nameref in the job-pool helpers.
+ # shellcheck disable=SC2034
+ image_job_labels["$!"]="image job for photo $photo"
done < <(incoming_image_files)
- wait_for_image_jobs image_job_pids failed
+ wait_for_image_jobs \
+ image_job_pids \
+ image_job_statuses \
+ image_job_labels \
+ failed
if (( failed != 0 )); then
return 1
fi
diff --git a/src/lib/job-pool.source.sh b/src/lib/job-pool.source.sh
index 5c629aa..a50a6e8 100644
--- a/src/lib/job-pool.source.sh
+++ b/src/lib/job-pool.source.sh
@@ -1,13 +1,25 @@
wait_for_parallel_job_pid() {
local -r pid="$1"; shift
- local -i status=0
+ local -r job_statuses_name="$1"; shift
+ local -n status_ref="$1"; shift
+ # shellcheck disable=SC2178
+ local -n job_statuses_ref="$job_statuses_name"
+ local -i wait_status=0
+
+ if [ -n "${job_statuses_ref[$pid]+x}" ]; then
+ # shellcheck disable=SC2034
+ status_ref="${job_statuses_ref[$pid]}"
+ unset "job_statuses_ref[$pid]"
+ return
+ fi
set +e
wait "$pid"
- status=$?
+ wait_status=$?
set -e
- return "$status"
+ # shellcheck disable=SC2034
+ status_ref="$wait_status"
}
parallel_job_pid_is_running() {
@@ -23,13 +35,42 @@ parallel_job_pid_is_running() {
return 1
}
+log_parallel_job_failure() {
+ local -r pid="$1"; shift
+ local -r status="$1"; shift
+ local -r job_labels_name="$1"; shift
+ # shellcheck disable=SC2178
+ local -n job_labels_ref="$job_labels_name"
+ local label
+
+ label="${job_labels_ref[$pid]:-pid $pid}"
+ printf 'ERROR: parallel job failed (%s): %s\n' "$status" "$label" >&2
+}
+
+finish_parallel_job() {
+ local -r pid="$1"; shift
+ local -r status="$1"; shift
+ local -r job_labels_name="$1"; shift
+ local -n failed_target_ref="$1"; shift
+ # shellcheck disable=SC2178
+ local -n job_labels_ref="$job_labels_name"
+
+ if (( status != 0 )); then
+ # shellcheck disable=SC2034
+ failed_target_ref=1
+ log_parallel_job_failure "$pid" "$status" "$job_labels_name"
+ fi
+
+ unset "job_labels_ref[$pid]"
+}
+
reap_finished_parallel_jobs() {
local -r job_pids_name="$1"; shift
+ local -r job_statuses_name="$1"; shift
+ local -r job_labels_name="$1"; shift
# shellcheck disable=SC2178
local -n job_pids_ref="$job_pids_name"
local -n failed_ref="$1"; shift
- local -i reaped_status="$1"; shift
- local -i has_reaped_status=1
local -i status=0
local -a remaining_pids=()
local -a running_pids=()
@@ -43,19 +84,8 @@ reap_finished_parallel_jobs() {
continue
fi
- if wait_for_parallel_job_pid "$pid"; then
- status=0
- else
- status=$?
- fi
- if (( status == 127 && has_reaped_status != 0 )); then
- status=$reaped_status
- has_reaped_status=0
- fi
-
- if (( status != 0 )); then
- failed_ref=1
- fi
+ wait_for_parallel_job_pid "$pid" "$job_statuses_name" status
+ finish_parallel_job "$pid" "$status" "$job_labels_name" failed_ref
done
job_pids_ref=("${remaining_pids[@]}")
@@ -63,75 +93,123 @@ reap_finished_parallel_jobs() {
wait_for_next_parallel_job() {
local -r job_pids_name="$1"; shift
+ local -r job_statuses_name="$1"; shift
+ local -r job_labels_name="$1"; shift
local -r failed_name="$1"; shift
# shellcheck disable=SC2178
local -n job_pids_ref="$job_pids_name"
+ # shellcheck disable=SC2178
+ local -n job_statuses_ref="$job_statuses_name"
+ local completed_pid=''
local -i status=0
set +e
- wait -n "${job_pids_ref[@]}"
+ wait -n -p completed_pid "${job_pids_ref[@]}"
status=$?
set -e
- reap_finished_parallel_jobs "$job_pids_name" "$failed_name" "$status"
+ if [ -n "$completed_pid" ]; then
+ job_statuses_ref["$completed_pid"]="$status"
+ fi
+
+ reap_finished_parallel_jobs \
+ "$job_pids_name" \
+ "$job_statuses_name" \
+ "$job_labels_name" \
+ "$failed_name"
}
wait_for_parallel_job_slot() {
local -r job_pids_name="$1"; shift
+ local -r job_statuses_name="$1"; shift
+ local -r job_labels_name="$1"; shift
local -r failed_name="$1"; shift
local -r max_jobs="$1"; shift
# shellcheck disable=SC2178
local -n job_pids_ref="$job_pids_name"
while (( ${#job_pids_ref[@]} >= max_jobs )); do
- wait_for_next_parallel_job "$job_pids_name" "$failed_name"
+ wait_for_next_parallel_job \
+ "$job_pids_name" \
+ "$job_statuses_name" \
+ "$job_labels_name" \
+ "$failed_name"
done
}
wait_for_parallel_jobs() {
# shellcheck disable=SC2178
local -n job_pids_ref="$1"; shift
+ local -r job_statuses_name="$1"; shift
+ local -r job_labels_name="$1"; shift
local -n failed_ref="$1"; shift
+ local -i status=0
: "$failed_ref"
while (( ${#job_pids_ref[@]} > 0 )); do
- if ! wait_for_parallel_job_pid "${job_pids_ref[0]}"; then
- failed_ref=1
- fi
+ wait_for_parallel_job_pid \
+ "${job_pids_ref[0]}" \
+ "$job_statuses_name" \
+ status
+ finish_parallel_job \
+ "${job_pids_ref[0]}" \
+ "$status" \
+ "$job_labels_name" \
+ failed_ref
job_pids_ref=("${job_pids_ref[@]:1}")
done
}
wait_for_image_job_slot() {
local -r image_job_pids_name="$1"; shift
+ local -r image_job_statuses_name="$1"; shift
+ local -r image_job_labels_name="$1"; shift
local -r failed_name="$1"; shift
wait_for_parallel_job_slot \
"$image_job_pids_name" \
+ "$image_job_statuses_name" \
+ "$image_job_labels_name" \
"$failed_name" \
"$IMAGE_JOBS"
}
wait_for_image_jobs() {
local -r image_job_pids_name="$1"; shift
+ local -r image_job_statuses_name="$1"; shift
+ local -r image_job_labels_name="$1"; shift
local -r failed_name="$1"; shift
- wait_for_parallel_jobs "$image_job_pids_name" "$failed_name"
+ wait_for_parallel_jobs \
+ "$image_job_pids_name" \
+ "$image_job_statuses_name" \
+ "$image_job_labels_name" \
+ "$failed_name"
}
wait_for_template_render_job_slot() {
local -r render_job_pids_name="$1"; shift
+ local -r render_job_statuses_name="$1"; shift
+ local -r render_job_labels_name="$1"; shift
local -r failed_name="$1"; shift
wait_for_parallel_job_slot \
"$render_job_pids_name" \
+ "$render_job_statuses_name" \
+ "$render_job_labels_name" \
"$failed_name" \
"$IMAGE_JOBS"
}
wait_for_template_render_jobs() {
local -r render_job_pids_name="$1"; shift
+ local -r render_job_statuses_name="$1"; shift
+ local -r render_job_labels_name="$1"; shift
local -r failed_name="$1"; shift
- wait_for_parallel_jobs "$render_job_pids_name" "$failed_name"
+ wait_for_parallel_jobs \
+ "$render_job_pids_name" \
+ "$render_job_statuses_name" \
+ "$render_job_labels_name" \
+ "$failed_name"
}