Odoo Performance Optimization: Speed Up Your Server

"Odoo is slow" is subjective. Measure first:

Response time benchmarks:

• Homepage load: < 2 seconds (logged-in user, no cache)
• Sales order creation: < 3 seconds
• Invoice PDF generation: < 5 seconds
• Product search (500 products): < 1 second
• Report generation (100-line table): < 10 seconds

If your Odoo exceeds these times consistently, performance tuning is needed.

Built-in profiling:
Enable Odoo's built-in profiler in odoo.conf:
limit_time_real = 300
limit_time_cpu = 180
log_level = debug

Slow queries (> 5 seconds) will be logged to /var/log/odoo/odoo-server.log with full SQL and stack trace.

Odoo uses multi-process workers to handle concurrent requests. Too few workers = slow response under load. Too many = RAM exhaustion and context-switching overhead.

Worker formula (for production):
workers = (CPU_cores * 2) + 1

For a 4-core server: workers = 9

In odoo.conf:
workers = 9
max_cron_threads = 2
limit_memory_hard = 2684354560 # 2.5 GB per worker
limit_memory_soft = 2147483648 # 2 GB

Restart Odoo:
sudo systemctl restart odoo

Monitor RAM usage:
htop

Each worker can consume 1-2 GB under load. For 9 workers, you need at least 16 GB RAM. If workers frequently hit the memory limit and restart, reduce the worker count or upgrade RAM.

PostgreSQL's shared_buffers cache frequently accessed data in RAM, reducing disk I/O.

Edit /etc/postgresql/*/main/postgresql.conf:
shared_buffers = 4GB # 25% of total RAM
effective_cache_size = 12GB # 75% of total RAM
work_mem = 64MB
maintenance_work_mem = 512MB

Restart PostgreSQL:
sudo systemctl restart postgresql

Verify settings:
sudo -u postgres psql
SHOW shared_buffers;
SHOW effective_cache_size;

If queries are still slow, enable query statistics:
CREATE EXTENSION IF NOT EXISTS pg_stat_statements;

Then query the slowest 10:
SELECT query, calls, total_time, mean_time
FROM pg_stat_statements
ORDER BY total_time DESC
LIMIT 10;

Missing indexes are the #1 cause of slow Odoo queries. A table scan on 100,000 records takes seconds; an indexed lookup takes milliseconds.

Identify missing indexes:
Odoo logs slow queries with EXPLAIN output when log_level = debug. Look for "Seq Scan" (sequential scan) in the logs — these are table scans that need an index.

Common indexes to add:
CREATE INDEX idx_sale_order_partner ON sale_order(partner_id);
CREATE INDEX idx_account_move_date ON account_move(date);
CREATE INDEX idx_stock_move_product ON stock_move(product_id);
CREATE INDEX idx_res_partner_name ON res_partner(name);

After adding indexes, analyze tables:
ANALYZE;

This updates PostgreSQL's query planner statistics so it uses the new indexes.

Odoo serves CSS, JS, and images on every request. Without caching, browsers re-download these files unnecessarily.

Nginx caching configuration:
location ~* \.(css|js|png|jpg|jpeg|gif|ico|woff|woff2|ttf)$ {
expires 30d;
add_header Cache-Control "public, immutable";
proxy_pass http://127.0.0.1:8069;
}

Reload Nginx:
sudo systemctl reload nginx

Test caching:
curl -I https://odoo.yourcompany.com/web/static/src/css/bootstrap.min.css

You should see:
Cache-Control: public, immutable
Expires: [30 days from now]

Browsers will cache these files for 30 days, reducing page load time from 3 seconds to under 1 second.

If your users are geographically distributed, serving assets from a single server adds latency. A CDN (Content Delivery Network) caches assets at edge locations worldwide.

Cloudflare (free tier):
1. Sign up at cloudflare.com
2. Add your domain and update nameservers
3. Enable "Cache Level: Standard" in Cloudflare dashboard
4. Set Browser Cache TTL to "1 month"

Cloudflare automatically caches /web/static/* and serves it from the nearest edge location. Users in Europe get assets from European servers; users in Asia get them from Asian servers.

Large tables slow down queries even with indexes. A sale_order table with 500,000 rows takes longer to search than one with 50,000.

Identify large tables:
sudo -u postgres psql odoo_production
SELECT schemaname, tablename, pg_size_pretty(pg_total_relation_size(schemaname||'.'||tablename)) AS size
FROM pg_tables
WHERE schemaname NOT IN ('pg_catalog', 'information_schema')
ORDER BY pg_total_relation_size(schemaname||'.'||tablename) DESC
LIMIT 10;

Archive old data:
For records older than 2 years that you rarely access:

• Export them to a separate "archive" database
• Delete from production
• Keep the archive database offline except when needed for historical lookups

This keeps the production database small and fast.

Reports, data imports, and month-end closing operations consume CPU and lock database tables. Running them during business hours degrades performance for everyone.

Move heavy cron jobs to night:
In Odoo, go to Settings → Technical → Automation → Scheduled Actions
Set "Next Execution Date" to 2:00 AM for:

• Inventory valuation
• Aged receivables report
• Data cleanup (expired sessions, old logs)

Database maintenance:
Schedule VACUUM and ANALYZE in a cron job at 3 AM:
0 3 * * * /usr/bin/sudo -u postgres psql odoo_production -c "VACUUM ANALYZE;"

This reclaims disk space and updates query planner statistics without user impact.