feat(workforce): add ClickHouse workforce analytics contour

This commit is contained in:
igor04091968
2026-06-23 21:52:57 +03:00
parent 341799ff9d
commit 337da4bb5f
44 changed files with 3973 additions and 461 deletions
+169
View File
@@ -0,0 +1,169 @@
#!/usr/bin/env bash
set -euo pipefail
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
ROOT_DIR="$(cd "$SCRIPT_DIR/.." && pwd)"
CATALOG_DIR="${CATALOG_DIR:-$ROOT_DIR/catalog}"
CLICKHOUSE_CONTAINER="${CLICKHOUSE_CONTAINER:-aw-rus-workforce-clickhouse}"
CLICKHOUSE_DATABASE="${CLICKHOUSE_DATABASE:-aw_workforce}"
CLICKHOUSE_USER="${CLICKHOUSE_USER:-}"
CLICKHOUSE_PASSWORD="${CLICKHOUSE_PASSWORD:-}"
CLICKHOUSE_CLIENT_BIN="${CLICKHOUSE_CLIENT_BIN:-clickhouse-client}"
CLICKHOUSE_READY_TIMEOUT_SEC="${CLICKHOUSE_READY_TIMEOUT_SEC:-60}"
REBUILD_AGGREGATES="${REBUILD_AGGREGATES:-0}"
client_auth_args=()
if [[ -n "$CLICKHOUSE_USER" ]]; then
client_auth_args+=(--user "$CLICKHOUSE_USER")
fi
if [[ -n "$CLICKHOUSE_PASSWORD" ]]; then
client_auth_args+=(--password "$CLICKHOUSE_PASSWORD")
fi
run_query() {
local query="$1"
if docker ps --format '{{.Names}}' | grep -Fxq "$CLICKHOUSE_CONTAINER"; then
docker exec -i "$CLICKHOUSE_CONTAINER" clickhouse-client "${client_auth_args[@]}" \
--database "$CLICKHOUSE_DATABASE" --query "$query"
return
fi
if command -v "$CLICKHOUSE_CLIENT_BIN" >/dev/null 2>&1; then
"$CLICKHOUSE_CLIENT_BIN" "${client_auth_args[@]}" \
--database "$CLICKHOUSE_DATABASE" --query "$query"
return
fi
printf 'No running ClickHouse container "%s" and no %s in PATH\n' \
"$CLICKHOUSE_CONTAINER" "$CLICKHOUSE_CLIENT_BIN" >&2
return 127
}
run_query_file() {
local query_file="$1"
if docker ps --format '{{.Names}}' | grep -Fxq "$CLICKHOUSE_CONTAINER"; then
docker exec -i "$CLICKHOUSE_CONTAINER" clickhouse-client "${client_auth_args[@]}" \
--multiquery <"$query_file"
return
fi
if command -v "$CLICKHOUSE_CLIENT_BIN" >/dev/null 2>&1; then
"$CLICKHOUSE_CLIENT_BIN" "${client_auth_args[@]}" --multiquery <"$query_file"
return
fi
printf 'No running ClickHouse container "%s" and no %s in PATH\n' \
"$CLICKHOUSE_CONTAINER" "$CLICKHOUSE_CLIENT_BIN" >&2
return 127
}
run_insert_file() {
local query="$1"
local data_file="$2"
if docker ps --format '{{.Names}}' | grep -Fxq "$CLICKHOUSE_CONTAINER"; then
docker exec -i "$CLICKHOUSE_CONTAINER" clickhouse-client "${client_auth_args[@]}" \
--database "$CLICKHOUSE_DATABASE" --query "$query" <"$data_file"
return
fi
if command -v "$CLICKHOUSE_CLIENT_BIN" >/dev/null 2>&1; then
"$CLICKHOUSE_CLIENT_BIN" "${client_auth_args[@]}" \
--database "$CLICKHOUSE_DATABASE" --query "$query" <"$data_file"
return
fi
printf 'No running ClickHouse container "%s" and no %s in PATH\n' \
"$CLICKHOUSE_CONTAINER" "$CLICKHOUSE_CLIENT_BIN" >&2
return 127
}
wait_for_clickhouse() {
local deadline
deadline=$((SECONDS + CLICKHOUSE_READY_TIMEOUT_SEC))
while (( SECONDS < deadline )); do
if run_query "SELECT 1" >/dev/null 2>&1; then
return 0
fi
sleep 1
done
printf 'ClickHouse is not ready after %s seconds\n' "$CLICKHOUSE_READY_TIMEOUT_SEC" >&2
return 1
}
require_file() {
local path="$1"
if [[ ! -f "$path" ]]; then
printf 'Required catalog file is missing: %s\n' "$path" >&2
return 1
fi
}
wait_for_clickhouse
require_file "$CATALOG_DIR/workstation_users.tsv"
require_file "$CATALOG_DIR/application_categories.tsv"
require_file "$CATALOG_DIR/domain_categories.tsv"
printf '[catalog] truncate dimension tables\n'
run_query "TRUNCATE TABLE $CLICKHOUSE_DATABASE.dim_workstation_user"
run_query "TRUNCATE TABLE $CLICKHOUSE_DATABASE.dim_application_category"
run_query "TRUNCATE TABLE $CLICKHOUSE_DATABASE.dim_domain_category"
printf '[catalog] load workstation users\n'
run_insert_file "
INSERT INTO $CLICKHOUSE_DATABASE.dim_workstation_user
(host_name, user_login, user_domain, employee_id, employee_name, department, branch, position, source, is_active)
FORMAT TabSeparatedWithNames
" "$CATALOG_DIR/workstation_users.tsv"
printf '[catalog] load application categories\n'
run_insert_file "
INSERT INTO $CLICKHOUSE_DATABASE.dim_application_category
(process_name, application_name, vendor, category, productivity_class, risk_level, is_system, source, comment, is_active)
FORMAT TabSeparatedWithNames
" "$CATALOG_DIR/application_categories.tsv"
printf '[catalog] load domain categories\n'
run_insert_file "
INSERT INTO $CLICKHOUSE_DATABASE.dim_domain_category
(domain, site_name, category, productivity_class, risk_level, business_allowed, source, comment, is_active)
FORMAT TabSeparatedWithNames
" "$CATALOG_DIR/domain_categories.tsv"
printf '[catalog] reload dictionaries\n'
run_query "SYSTEM RELOAD DICTIONARY $CLICKHOUSE_DATABASE.dict_workstation_user"
run_query "SYSTEM RELOAD DICTIONARY $CLICKHOUSE_DATABASE.dict_application_category"
run_query "SYSTEM RELOAD DICTIONARY $CLICKHOUSE_DATABASE.dict_domain_category"
if [[ "$REBUILD_AGGREGATES" == "1" ]]; then
printf '[catalog] rebuild aggregates\n'
run_query_file "$ROOT_DIR/admin/rebuild_aggregates.sql"
fi
printf '[catalog] dictionary status\n'
run_query "
SELECT name, status, last_exception
FROM system.dictionaries
WHERE database = '$CLICKHOUSE_DATABASE'
AND name IN ('dict_workstation_user', 'dict_application_category', 'dict_domain_category')
ORDER BY name
FORMAT PrettyCompact
"
printf '[catalog] raw unknown summary\n'
run_query "
SELECT 'subjects' AS area, count() AS rows
FROM $CLICKHOUSE_DATABASE.v_workforce_unknown_subjects
UNION ALL
SELECT 'processes' AS area, count() AS rows
FROM $CLICKHOUSE_DATABASE.v_workforce_unknown_processes
UNION ALL
SELECT 'domains' AS area, count() AS rows
FROM $CLICKHOUSE_DATABASE.v_workforce_unknown_domains
FORMAT PrettyCompact
"
@@ -0,0 +1,14 @@
AW_WORKFORCE_AW_URL=http://10.10.10.13:5600/api/0
AW_WORKFORCE_CLICKHOUSE_URL=http://10.10.10.2:8123
AW_WORKFORCE_CLICKHOUSE_DATABASE=aw_workforce
AW_WORKFORCE_CLICKHOUSE_USER=
AW_WORKFORCE_CLICKHOUSE_PASSWORD=
AW_WORKFORCE_HOST=SHARKON2025
AW_WORKFORCE_STATE_PATH=/var/lib/aw-workforce-ingest/state.json
AW_WORKFORCE_OVERLAP_SECONDS=300
AW_WORKFORCE_RETRY_ATTEMPTS=3
AW_WORKFORCE_RETRY_BACKOFF_MS=1000
AW_WORKFORCE_FAIL_ON_EMPTY=false
AW_WORKFORCE_JSON=true
no_proxy=localhost,127.0.0.1,10.10.10.13,10.10.10.2,10.10.10.0/24
NO_PROXY=localhost,127.0.0.1,10.10.10.13,10.10.10.2,10.10.10.0/24
@@ -0,0 +1,17 @@
[Unit]
Description=AW-rus workforce ClickHouse ingest
After=network-online.target docker.service
Wants=network-online.target
[Service]
Type=oneshot
WorkingDirectory=/opt/activitywatch/clickhouse-workforce
EnvironmentFile=-/etc/activitywatch/aw-workforce-ingest.env
ExecStart=/usr/local/bin/aw-workforce-ingest
TimeoutStartSec=20min
User=root
Group=root
Nice=5
StandardOutput=journal
StandardError=journal
SyslogIdentifier=aw-workforce-ingest
@@ -0,0 +1,13 @@
[Unit]
Description=Run AW-rus workforce ClickHouse ingest every 5 minutes
[Timer]
OnBootSec=2min
OnUnitActiveSec=5min
AccuracySec=30s
RandomizedDelaySec=30s
Persistent=true
Unit=aw-workforce-ingest.service
[Install]
WantedBy=timers.target
@@ -0,0 +1,34 @@
#!/usr/bin/env bash
set -euo pipefail
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
ROOT="${AW_WORKFORCE_ROOT:-/opt/activitywatch/clickhouse-workforce}"
ENV_DIR="${AW_WORKFORCE_ENV_DIR:-/etc/activitywatch}"
STATE_DIR="${AW_WORKFORCE_STATE_DIR:-/var/lib/aw-workforce-ingest}"
SYSTEMD_DIR="${AW_WORKFORCE_SYSTEMD_DIR:-/etc/systemd/system}"
install -d -m 0755 "$ROOT" "$ENV_DIR" "$STATE_DIR" "$SYSTEMD_DIR"
if [[ ! -f "$ENV_DIR/aw-workforce-ingest.env" ]]; then
install -m 0640 "$SCRIPT_DIR/aw-workforce-ingest.env.example" \
"$ENV_DIR/aw-workforce-ingest.env"
fi
install -m 0644 "$SCRIPT_DIR/aw-workforce-ingest.service" \
"$SYSTEMD_DIR/aw-workforce-ingest.service"
install -m 0644 "$SCRIPT_DIR/aw-workforce-ingest.timer" \
"$SYSTEMD_DIR/aw-workforce-ingest.timer"
if command -v systemctl >/dev/null 2>&1; then
systemctl daemon-reload
fi
cat <<EOF
Installed aw-workforce-ingest runtime files.
Next manual deployment steps:
install -m 0755 <built aw-workforce-ingest binary> /usr/local/bin/aw-workforce-ingest
edit $ENV_DIR/aw-workforce-ingest.env
systemctl enable --now aw-workforce-ingest.timer
systemctl start aw-workforce-ingest.service
EOF
@@ -0,0 +1,48 @@
#!/usr/bin/env bash
set -euo pipefail
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
ROOT_DIR="$(cd "$SCRIPT_DIR/.." && pwd)"
CLICKHOUSE_CONTAINER="${CLICKHOUSE_CONTAINER:-aw-rus-workforce-clickhouse}"
CLICKHOUSE_DATABASE="${CLICKHOUSE_DATABASE:-aw_workforce}"
CLICKHOUSE_CLIENT_BIN="${CLICKHOUSE_CLIENT_BIN:-clickhouse-client}"
LIMIT="${LIMIT:-50}"
run_query() {
local query="$1"
if docker ps --format '{{.Names}}' | grep -Fxq "$CLICKHOUSE_CONTAINER"; then
docker exec -i "$CLICKHOUSE_CONTAINER" clickhouse-client \
--database "$CLICKHOUSE_DATABASE" --query "$query"
return
fi
"$CLICKHOUSE_CLIENT_BIN" --database "$CLICKHOUSE_DATABASE" --query "$query"
}
cd "$ROOT_DIR"
printf '\n[unknown subjects]\n'
run_query "
SELECT *
FROM $CLICKHOUSE_DATABASE.v_workforce_unknown_subjects
LIMIT $LIMIT
FORMAT PrettyCompact
"
printf '\n[unknown processes]\n'
run_query "
SELECT *
FROM $CLICKHOUSE_DATABASE.v_workforce_unknown_processes
LIMIT $LIMIT
FORMAT PrettyCompact
"
printf '\n[unknown domains]\n'
run_query "
SELECT *
FROM $CLICKHOUSE_DATABASE.v_workforce_unknown_domains
LIMIT $LIMIT
FORMAT PrettyCompact
"
+136
View File
@@ -0,0 +1,136 @@
#!/usr/bin/env bash
set -euo pipefail
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
ROOT_DIR="$(cd "$SCRIPT_DIR/.." && pwd)"
CLICKHOUSE_CONTAINER="${CLICKHOUSE_CONTAINER:-aw-rus-workforce-clickhouse}"
CLICKHOUSE_DATABASE="${CLICKHOUSE_DATABASE:-aw_workforce}"
CLICKHOUSE_USER="${CLICKHOUSE_USER:-}"
CLICKHOUSE_PASSWORD="${CLICKHOUSE_PASSWORD:-}"
CLICKHOUSE_CLIENT_BIN="${CLICKHOUSE_CLIENT_BIN:-clickhouse-client}"
LOAD_DEMO_SEED="${LOAD_DEMO_SEED:-1}"
CLICKHOUSE_READY_TIMEOUT_SEC="${CLICKHOUSE_READY_TIMEOUT_SEC:-60}"
client_auth_args=()
if [[ -n "$CLICKHOUSE_USER" ]]; then
client_auth_args+=(--user "$CLICKHOUSE_USER")
fi
if [[ -n "$CLICKHOUSE_PASSWORD" ]]; then
client_auth_args+=(--password "$CLICKHOUSE_PASSWORD")
fi
run_client() {
local query_file="${1:-}"
local query="${2:-}"
if docker ps --format '{{.Names}}' | grep -Fxq "$CLICKHOUSE_CONTAINER"; then
if [[ -n "$query_file" ]]; then
docker exec -i "$CLICKHOUSE_CONTAINER" clickhouse-client "${client_auth_args[@]}" --multiquery <"$query_file"
else
docker exec -i "$CLICKHOUSE_CONTAINER" clickhouse-client "${client_auth_args[@]}" --database "$CLICKHOUSE_DATABASE" --query "$query"
fi
return
fi
if command -v "$CLICKHOUSE_CLIENT_BIN" >/dev/null 2>&1; then
if [[ -n "$query_file" ]]; then
"$CLICKHOUSE_CLIENT_BIN" "${client_auth_args[@]}" --multiquery <"$query_file"
else
"$CLICKHOUSE_CLIENT_BIN" "${client_auth_args[@]}" --database "$CLICKHOUSE_DATABASE" --query "$query"
fi
return
fi
printf 'No running ClickHouse container "%s" and no %s in PATH\n' \
"$CLICKHOUSE_CONTAINER" "$CLICKHOUSE_CLIENT_BIN" >&2
return 127
}
wait_for_clickhouse() {
local deadline
deadline=$((SECONDS + CLICKHOUSE_READY_TIMEOUT_SEC))
while (( SECONDS < deadline )); do
if run_client "" "SELECT 1" >/dev/null 2>&1; then
return 0
fi
sleep 1
done
printf 'ClickHouse is not ready after %s seconds\n' "$CLICKHOUSE_READY_TIMEOUT_SEC" >&2
return 1
}
apply_sql_dir() {
local sql_file
for sql_file in "$ROOT_DIR"/clickhouse/init/*.sql; do
printf '[sql] %s\n' "${sql_file#$ROOT_DIR/}"
run_client "$sql_file" ""
done
}
assert_scalar_nonzero() {
local name="$1"
local query="$2"
local value
value="$(run_client "" "$query" | tr -d '[:space:]')"
if [[ ! "$value" =~ ^[0-9]+$ ]] || (( value < 1 )); then
printf '[FAIL] %s: expected positive integer, got "%s"\n' "$name" "$value" >&2
return 1
fi
printf '[OK] %s: %s\n' "$name" "$value"
}
assert_no_dictionary_errors() {
local errors
errors="$(run_client "" "
SELECT count()
FROM system.dictionaries
WHERE database = '$CLICKHOUSE_DATABASE'
AND name IN ('dict_workstation_user', 'dict_application_category', 'dict_domain_category')
AND (status != 'LOADED' OR last_exception != '')
")"
errors="$(printf '%s' "$errors" | tr -d '[:space:]')"
if [[ "$errors" != "0" ]]; then
printf '[FAIL] dictionaries have load errors\n' >&2
run_client "" "
SELECT database, name, status, last_exception
FROM system.dictionaries
WHERE database = '$CLICKHOUSE_DATABASE'
AND name IN ('dict_workstation_user', 'dict_application_category', 'dict_domain_category')
FORMAT Vertical
"
return 1
fi
printf '[OK] dictionaries loaded\n'
}
wait_for_clickhouse
apply_sql_dir
if [[ "$LOAD_DEMO_SEED" == "1" ]]; then
printf '[sql] sample/seed_demo.sql\n'
run_client "$ROOT_DIR/sample/seed_demo.sql" ""
fi
assert_no_dictionary_errors
assert_scalar_nonzero "dictionary count" "
SELECT count()
FROM system.dictionaries
WHERE database = '$CLICKHOUSE_DATABASE'
AND name IN ('dict_workstation_user', 'dict_application_category', 'dict_domain_category')
"
assert_scalar_nonzero "hourly aggregate rows" "
SELECT count()
FROM $CLICKHOUSE_DATABASE.agg_workforce_productivity_hourly
"
assert_scalar_nonzero "daily productivity rows" "
SELECT count()
FROM $CLICKHOUSE_DATABASE.v_workforce_productivity_daily
"
assert_scalar_nonzero "unknown quality rows" "
SELECT count()
FROM $CLICKHOUSE_DATABASE.v_workforce_unknown_quality_daily
"
printf '[OK] ClickHouse workforce smoke completed\n'