umask for data

[kronwerk]

data files (both AOF and RDB) are created now using hardcoded filemode (0644).
that can be inconvenient in some cases.
in this PR it's proposed to create that files using server.umask providing more flexibility for a user.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 71.04%. Comparing base (0cc0bf7) to head (8459f02).

Additional details and impacted files

@@             Coverage Diff              @@
##           unstable    #1725      +/-   ##
============================================
+ Coverage     70.87%   71.04%   +0.16%     
============================================
  Files           123      123              
  Lines         65651    65651              
============================================
+ Hits          46529    46640     +111     
+ Misses        19122    19011     -111     

Files with missing lines	Coverage Δ
src/aof.c	80.24% <100.00%> (ø)
src/replication.c	87.14% <100.00%> (-0.13%)	⬇️

... and 11 files with indirect coverage changes

[rjd15372]

I don't have much experience in administrating valkey deployments, but maybe using the server binary mask isn't also what we're expecting.
If 644 is not flexible, then maybe we need a config option to override the default file permissions for both the rdb and aof files.

[kronwerk]

I don't have much experience in administrating valkey deployments, but maybe using the server binary mask isn't also what we're expecting. If 644 is not flexible, then maybe we need a config option to override the default file permissions for both the rdb and aof files.

yeah, that was another option - but it appears to be an overkill to create 1 (or may be even 2, for AOF and RDB separately) another config setting to achieve this flexibility. I (maybe I'm wrong?) can't find a situation there we gonna need different permissions for AOF and RDB - and Valkey binary and AOF|RDB also

[kronwerk]

@madolson hi, could you please take a look at this PR? thx