DSpace Dashboard

Web dashboard for DSpace: summary stats, downloads, submitters, and ORCID sync views with admin-only access.

Short guide to run the app, configure environment, and install a systemd unit.

Requirements

• Python 3.9+
• Access to DSpace REST API, Solr, and PostgreSQL

Quick start (manual)

cd /opt/dspace-dashboard
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

Environment file (/etc/default/dspace-dashboard)

Create the environment file used by systemd:

sudo tee /etc/default/dspace-dashboard >/dev/null <<'EOF'
# Path to DSpace local.cfg
DSPACE_CONFIG_PATH=/dspace/config/local.cfg

# Flask session secret
SECRET_KEY=replace_with_random_64_chars

# Optional settings
CACHE_TTL_SECONDS=300
START_YEAR=2025
START_MONTH=1

# ORCID metadata field id (required, varies by DSpace instance)
ORCID_FIELD_ID=205
EOF

Systemd unit

Create the unit file:

sudo tee /etc/systemd/system/dspace-dashboard.service >/dev/null <<'EOF'
[Unit]
Description=DSpace Dashboard (Flask via Gunicorn)
After=network-online.target
Wants=network-online.target

[Service]
Type=simple
User=dspace
Group=dspace
WorkingDirectory=/opt/dspace-dashboard

EnvironmentFile=/etc/default/dspace-dashboard

# Gunicorn слушает localhost или 0.0.0.0
ExecStart=/opt/dspace-dashboard/venv/bin/gunicorn \
  --workers 2 \
  --threads 4 \
  --timeout 60 \
  --bind localhost:8088 \
  app:app

Restart=always
RestartSec=3

[Install]
WantedBy=multi-user.target
EOF

Enable and start:

sudo systemctl daemon-reload
sudo systemctl enable --now dspace-dashboard

Open http://localhost:8088 (or your reverse-proxy URL).

Logs:

sudo journalctl -u dspace-dashboard -f

Item edits from DSpace logs

Dashboard section "Редагування" uses events parsed from DSpace logs.

Parser daemon (manual run)

Use the long-running daemon for near-real-time updates:

cd /opt/dspace-dashboard
source .venv/bin/activate
python3 parser_daemon.py --log-glob "/dspace/log/dspace*.log" --log-level INFO

Optional environment overrides

Set these in /etc/default/dspace-dashboard (used by both the dashboard and parser services) to fine-tune behavior:

• DSPACE_EDIT_LOG_GLOB — glob for log files (default /dspace/log/dspace.log).
• DSPACE_EDIT_POLL_SECONDS — sleep between iterations (default 5).
• DSPACE_EDIT_PENDING_SECONDS — delay before confirming edit (default 180).
• DSPACE_EDIT_DEDUPE_SECONDS — dedupe window for the same user+item edits (default 60).
• DSPACE_SYSTEM_EVENT_RETENTION_HOURS — how long to keep workflow/system markers (default 48).
• DSPACE_REQUEST_CONTEXT_RETENTION_SECONDS — in-memory request context TTL (default 900).
• DSPACE_EDIT_LOG_LEVEL — log level for the daemon (INFO, DEBUG, etc.).

Systemd unit for parser

Create a companion unit that runs alongside the web app:

sudo tee /etc/systemd/system/dspace-dashboard-parser.service >/dev/null <<'EOF'
[Unit]
Description=DSpace Dashboard Item Edit Parser
After=network-online.target postgresql.service
Wants=network-online.target

[Service]
Type=simple
User=dspace
Group=dspace
WorkingDirectory=/opt/dspace-dashboard
EnvironmentFile=/etc/default/dspace-dashboard
ExecStart=/opt/dspace-dashboard/.venv/bin/python3 parser_daemon.py \
  --log-glob "/dspace/log/dspace*.log"
Restart=always
RestartSec=5

[Install]
WantedBy=multi-user.target
EOF

Enable and start the daemon (after deploying code and migrations):

sudo systemctl daemon-reload
sudo systemctl enable --now dspace-dashboard-parser

Follow logs via sudo journalctl -u dspace-dashboard-parser -f.

Notes:

• daemon keeps parser state in PostgreSQL (inode + offset) and resumes automatically;
• only confirmed update events are counted (ItemServiceImpl ::update_item:item_id=...);
• after the first daemon run, refresh the dashboard page /item-edits to populate charts.