mount: raise the listing cache default to a million children

A ceiling rather than a reservation: nothing is held for a directory
never walked, so this only materialises for a mount that really does walk
that many cached children, and one that does wants the speedup. A
chunkless child measured about 480 bytes, so the doc and the flag help
both say what reaching the ceiling costs -- near 450MB -- since that is
the number an operator needs to lower it against.

weed/command/mount.go

@@ -117,7 +117,7 @@ func init() {
 	mountOptions.cacheDirForWrite = cmdMount.Flag.String("cacheDirWrite", "", "buffer writes mostly for large files")
 	mountOptions.writeBufferSizeMB = cmdMount.Flag.Int64("writeBufferSizeMB", 0, "global cap on the per-mount write buffer (memory + swap) in MB, 0 means unlimited. Bounds /tmp growth when volume uploads stall")
 	mountOptions.cacheMetaTtlSec = cmdMount.Flag.Int("cacheMetaTtlSec", 60, "metadata cache validity seconds")
-	mountOptions.dirListCacheEntries = cmdMount.Flag.Int("dirListCacheEntries", meta_cache.DefaultListingCacheEntries, "directory children kept decoded in memory, so reopening a folder rereads nothing; 0 disables")
+	mountOptions.dirListCacheEntries = cmdMount.Flag.Int("dirListCacheEntries", meta_cache.DefaultListingCacheEntries, "directory children kept decoded in memory, so reopening a folder rereads nothing. A ceiling, not a reservation: only walked directories are held, at roughly half a kilobyte per child. 0 disables")
 	mountOptions.dataCenter = cmdMount.Flag.String("dataCenter", "", "prefer to write to the data center")
 	mountOptions.allowOthers = cmdMount.Flag.Bool("allowOthers", true, "allows other users to access the file system")
 	mountOptions.defaultPermissions = cmdMount.Flag.Bool("defaultPermissions", true, "enforce permissions by the operating system")

weed/mount/meta_cache/listing_cache.go

@@ -9,9 +9,14 @@ import (
 )
 
 // DefaultListingCacheEntries is how many directory children the mount will hold
-// decoded in memory across every cached listing. A chunkless entry runs a few
-// hundred bytes, so this is tens of megabytes.
-const DefaultListingCacheEntries = 200000
+// decoded in memory across every cached listing.
+//
+// This is a ceiling, not a reservation: nothing is held for a directory that was
+// never walked, so a mount over ordinary directories never approaches it. A
+// chunkless child measured about 480 bytes, so reaching the default means
+// holding something near 450MB, and a mount that wants a tighter bound should
+// lower it rather than rely on never walking that far.
+const DefaultListingCacheEntries = 1000000
 
 // listingCache keeps directories that have been walked all the way through, so
 // walking one again costs no store reads and no decoding. It holds only what the