Over a period of time the redo log space requests and redo buffer allocation retries parameter continuosly increase for my database. Once i restart the database it becomes zero and again the value increases.
Is there any way to keep the value low?
Mine is an online database with lots of inserts happening.
Do you have a performance problem? Do you have evidence that those wait events are causing the performance problem? On the database I look after buffer busy waits are always in my top 5 wait events yet whenever I trace a performance issue I find that buffer busy waits contribute a tiny amount to the overall runtime duration of the session. I could try to eliminate those buffer busy waits, but the performance problem will still be there after I've finished. There's no point dealing with these if there are bigger fish to fry. Make sure you're dealing with the cause of a problem before worrying about it would be my advice.
"Do you have a performance problem? Do you have evidence that those wait events are causing the performance problem?"
Unless you've done some benchmarks, rounding up the nearest suspects is not a good idea.
And benchmarking your system at startup is meaningless.
Until you've got the system cranked up to represent a typical database profile you are probably wasting your time. Until that shared pool and buffer cache are in full swing then it just isn't worth looking at any stats at all.
I don't have as any major performance problem. Most of my tuning statistics are fine like Library cache hit ratio, Dictionary cache hit ratio and others. Only the redo log space requests is increasing with time.
Meanwhile i am increasing my log buffer paramater and monitoring how the things are going on.