Guides for SaaS and CRM owners
| Article | Topics | Published |
|---|---|---|
Queue backlog — check the states before adding workersA growing ready count does not mean you need more workers. Split ready, delayed, reserved and failed, and check the consumers you already pay for. 7 min read |
||
Redis cache stampede: fix it before the bigger boxWhen a Redis key expires, many requests can rebuild it at once. Count those rebuilds before you buy more RAM or CPU for a bigger box. 9 min read |
||
Site technical audit: a short list with evidenceA technical audit is a short list with evidence: one request path, one timed operation, the release route, and where data can leak between accounts. 9 min read |
||
Process memory leak: what to check before buying RAMThe process grows until the server kills it. Record the size before, at the peak and after one repeated action, and you will see if memory comes back. 9 min read |
||
MySQL slow log: what to read before you refactorRead the slow query log before you split anything: group the statements, count repeats and rows examined, and see if the service must change. 9 min read |
||
Slow CRM/SaaS reports: indexes and queues firstA report in your CRM or SaaS makes people wait. Time it, check indexes, queues and account boundaries, and see if the product needs replacing at all. 8 min read |
||
Slow site, lost leads: what to check before a redesignThe site feels slow and leads are slipping. Name the page people wait on, time it, and find the query behind it before you pay for a redesign. 9 min read |
