- Accidental DBA
- Posts
- What DBAs Want You to Know
As with all roles, people in that role know it better than those who have a different primary function. It is critical that different teams work together to make the relationship a good one—for the individuals as well as the organization.
Here are 5 things that DBAs would love to see improve:
1. Communication With Development Teams
DBAs regularly deal with developers making changes unexpectedly. Unplanned schema updates or sudden deployments can break things quickly. Application development is different in some ways than database development.
Solution: Establish a CI/CD process, as well as a less formal “conversation” between the teams for best practices on new or changing code.
2. Expectations from Management
Management sometimes expects instant solutions to performance issues without understanding the underlying complexity. Near-instant response can be done but no DBA can guarantee instant resolution. Enterprise level databases can be adversely affected by DNS, code, hardware, networking, etc.
Solution: Set clear, measurable SLAs and have defined processes to investigate issues. Bring details to the issue, not vague descriptions (“the database is down!”)
3. Documentation
Inherited databases often come with little to no documentation, leaving the DBA to reverse-engineer critical information. Some of this is the DBA’s responsibility. Other parts might be the architect, systems admin or the development team.
Solution: Make documentation part of the handoff process for new projects or upgrades. Every server, every database. Also, comments in ALL code, including changes.
4. Reliance on DBAs for Non-DBA Tasks
DBAs are occasionally asked to handle tasks like managing backups for third-party apps or troubleshooting general server issues. In some cases, this is not an issue but sometimes it can stretch the DBA too thin, risking slower response times to critical issues.
Solution: Have a clear definition of the DBA role and understand what level of bandwidth is free.
5. Hardware Resources
DBAs often battle performance issues caused by undersized hardware or storage limitations imposed by budget constraints. You will almost never hear a DBA say you have too much memory in the server.
Solution: Measure the performance and resource utilization and see where you are under-allocated. Re-check this periodically.
The Bottom Line
This is a partial list of things that can improve the organization’s data stability, employee satisfaction and overall functioning. It all boils down to communication, as with most relationships. Have a clear pathway for that communication to be open and well-received on both sides.
5 more things in next week’s edition!
If you need help, please reach out:
Get a FREE SQL Server Health Check from Dallas DBAs!
New Pocket DBA® clients get the first month FREE!
Join the Accidental SQL DBA group on LinkedIn – no fee, no spam
My Recent LinkedIn Posts
Interesting Stuff I Read This Week
Intel Announces Retirement of CEO Pat Gelsinger | Business Wire – 45 years in one place…wow!
Rare spade-toothed whale - Only seven specimens of the spade-toothed whale have been documented since the 1880s.
UnitedHealthcare CEO Brian Thompson Fatally Shot in Manhattan - Bloomberg – This is never the answer.
Bitcoin soars past $100K milestone | LinkedIn – big gains make me wary of big losses.
SQL tidBITs:
Memory is 1st, CPU is 2nd, Storage is 3rd.
SQL Server performance relies more on memory than anything else. Many queries only need one CPU, since the logic is simple. They live and die in memory since SQL Server doesn’t read/write directly to disk. This is why major cloud vendors now offer “Memory optimized” servers specifically for database usage.
Reply