Sybase DBA plans for analyzing & monitoring (Most of these questions are valid for other db engines, too.)

· Read and comment most important DBA documents.

· Are there any audit or other documents that we should read?

· What is current backup and recovery plan for ASE databases?

· What are most critical and important ASE environments and databases?

· What kind of environmets (OLTP, DW etc.) do we have?

· Do we have any hybrid needs? If yes, is ASE configured on fly?

· How do you describe your reporting system?

· Where and what kind of dumps are copied and how dumps are backuped themselves?

· What kind of dumps are taken and how often?

· How long does it take to recover from biggest database, if it fails at 02.10 PM?

· How many transaction dumps are needed for recovery in this case?

· Can you describe some basic recovery scenarious?

· First one can be failover (prod node fails), second case happens at 02.05 PM (biggest db) when static table is truncated (software canno’t be used) and third recovery scenario is complete database corrupt and db (biggest) is unused at 02.10 PM and it’s Monday. Sybase can not revover it and is now suggesting (03.30 PM) long running recovery option. How long does it take to recover from these scenarious? Are other db’s in use at time? How are db servives treated/moved back to mother node when failed node is fixed (case 1)?

· What else needs to be backuped? Any bcp’s?

· What other important DBA utilities (statistics etc.) do we have?

· If you haven’t got anything better to do, simulate and learn your production backup and utility model.

· Try to find time and study your ASE configurations.

· Get familiar with system and take basic listings from ASE, file system and operating system (crontab –l).

· Code, develop, test and use ASE DBA and monitoring & analyzing scripts. This is quite important. We must prepare. Needed wheels are usually nowadays already invented, you just need to find them.

· List all good and possible graphical DBA tools and evaluate them.

· What things do we need to monitor and analyze? Resource hogs (I/O and CPU).

· How can we see ASE bottlenecks?

· What is monitored and analyzed right now? It seems, quite a lot indeed.

· What ASE monitoring is needed for Patrol?

· Make plans for monitoring and analyzing your ASE environments and databases.

· Study MDA analysis and monitoring model and simulate it in your sandbox ASE.

· How can you get all needed listings from different ASE systems?

· Crosscheck different listings, if needed.

· Choose best graphical DBA tools.

· What are most difficult show stoppers in your ASE systems?

· What happens if centralized tape robot is broken for five weeks? Where do we move dumps (full + tran)?

· Study and learn all other trace and debug methods and flags.

· Order and read good Sybase DBA books and basic DBA manuals. If you don’t know ask collage or phone or email to Sybase tech. Study also what Sybase and other sites and forums offers.