[chart/redis_ha][BUG] Startup probe has bad defaults

[hpfmn]

Describe the bug

The newly introduced startup probe gives the redis less time to startup, leading to crash loops for servers that take longer to startup. The repeated crashes also lead to the PVC filling up because new temp-rdb files are constantly created.

The "old" liveness probes use these defaults:

    initialDelaySeconds: 30
    periodSeconds: 15
    timeoutSeconds: 15
    successThreshold: 1
    failureThreshold: 5

That means it waits 30+(5*15)=105 seconds that redis is allowed to need for startup. The new startup probe has these defaults:

    initialDelaySeconds: 5
    periodSeconds: 10
    timeoutSeconds: 15
    successThreshold: 1
    failureThreshold: 3

Leading to 5+(3*15)=50 seconds slag for the redis, which is less than what the liveness probe allows and the opposite of what a startup probe is made for. To quote kubernetes documentation on startup probes:

The solution is to set up a startup probe with the same command, HTTP or TCP check, with a failureThreshold * periodSeconds long enough to cover the worst case startup time.

To Reproduce
Steps to reproduce the behavior:

1. Have a redis that needs a bit longer to startup
2. Update to a version newer than 4.27.8 which introduces the new startup probe
3. Experience CrashLoops for first new node that is rolled over
4. See error

Expected behavior

Use at least the same defaults as for the liveness probe!