A cap on the catalog's NVMe size: above it only the tail and the stored ranges

After reset 24 the catalog grew ~3.7 GB an hour (77 GB, root 31 GB free, 600k
history ranges to go). CATALOG_MAX_NVME_GB (as voyager 1's
BACKFILL_MAX_NVME_GB): once the default tablespace is over it, leases and the
prefetcher take priority < 0 only (the tail, the stored ranges); measured once
a minute. Off when unset; 90 on homelab.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

tests/functional/catalog_space.spec.ts

@@ -0,0 +1,22 @@
+import { test } from '@japa/runner'
+import coordinatorConfig from '#config/coordinator'
+
+test.group('CatalogSpace', () => {
+  test('is full over its cap, and off without one', async ({ assert }) => {
+    const before = coordinatorConfig.catalogMaxNvmeGb
+    try {
+      const { default: off } = await import('#services/catalog_space')
+      coordinatorConfig.catalogMaxNvmeGb = 0
+      assert.isFalse(await off.full())
+      // A fresh measurement with a cap far under any database's size.
+      ;(off as any).checkedAt = 0
+      coordinatorConfig.catalogMaxNvmeGb = 0.000001
+      assert.isTrue(await off.full())
+      ;(off as any).checkedAt = 0
+      coordinatorConfig.catalogMaxNvmeGb = 1_000_000
+      assert.isFalse(await off.full())
+    } finally {
+      coordinatorConfig.catalogMaxNvmeGb = before
+    }
+  })
+})