[Cherry-Pick][Metrics] Support cpu-cache-block-num (#6390)

[Copilot]

Motivation

Cherry-pick PR #6390 to release/2.4 branch. Adds max_cpu_block_num metric to report CPU cache block capacity alongside existing GPU block metrics.

Modifications

Core changes:

• fastdeploy/metrics/metrics.py: Added max_cpu_block_num Gauge metric definition to MetricsManager and METRIC_CONFIGS
• fastdeploy/cache_manager/prefix_cache_manager.py: Set max_cpu_block_num metric in __init__() and update_cache_config() methods using self.num_cpu_blocks

Documentation:

• docs/online_serving/metrics.md: Added fastdeploy:max_cpu_block_num metric entry
• docs/zh/online_serving/metrics.md: Added metric entry and clarified max_gpu_block_num description to specify GPU blocks

Usage or Command

The new metric is automatically reported when FastDeploy starts with CPU cache enabled:

fastdeploy:max_cpu_block_num (Gauge) - Total number of CPU blocks initialized at startup

Accuracy Tests

Documentation-only change for metrics reporting. No impact on model outputs or computation.

Checklist

• Add at least a tag in the PR title.
• Format your code, run pre-commit before commit.
• Add unit tests. No new tests needed - cherry-pick of already-reviewed code from develop branch.
• Provide accuracy results. N/A - metrics reporting only, no model changes.
• If the current PR is submitting to the release branch, make sure the PR has been submitted to the develop branch, then cherry-pick it to the release branch with the [Cherry-Pick] PR tag.

Original prompt

@PaddlePaddle/FastDeploy/pull/6390 这个PR Cherry pick至release/2.4分支

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[paddle-bot]

Thanks for your contribution!