tpu-client-next: use 1s for quic keep alive and 10s for timeout

[KirillLykov]

This PR decreases constants controling keep alive interval and connection timeout on the client side. For keep alive, 1 sec is chosen because on the mainnet most of the validators use timeout 2sec and if we employ 45s timeout as it is currently, all our connections timeout constantly.

See also #9213

[alexpyattaev]

If the server timeouts are indeed 2s, we should send keepalive every 950ms or so, such that if one is lost the second one would not have to race against the timeout but actually have a margin. Otherwise LGTM (though my heart bleeds for the amount of ACK spam this will create).

[KirillLykov]

There are some validators that use 2sec for whatever reason. FD are using 10s, we will use 10s in 4.0 as well. So with time we should be able to increase this at least to 4sec.
But also users are using 1sec in their client to avoid seeing these timeouts.

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 82.5%. Comparing base (729712b) to head (ee9bd05).
⚠️ Report is 1 commits behind head on master.

Additional details and impacted files

@@           Coverage Diff           @@
##           master   #10012   +/-   ##
=======================================
  Coverage    82.5%    82.5%           
=======================================
  Files         844      844           
  Lines      316758   316758           
=======================================
+ Hits       261579   261581    +2     
+ Misses      55179    55177    -2     

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.