What is a reasonable upper limit to the rt_hash_table.

[Robin Holt]

We have a system with a very large amount of memory.  We are noticing
pauses of approximately 5 seconds every 10 minutes.  We tracked it down
to rt_run_flush holding off other timer processing while it scans the
rt_hash_table.  The following is from the boot:

IP: routing cache hash table of 33554432 buckets, 524288Kbytes

This seems like an outrageously large value.  I realize the 2.6 kernel
has rhash_entries as a boot option.

Can I get some guidance on what a reasonable upper limit would be?
What is this guidance based upon?

What is the reason for not making that upper limit a default and let
rhash_entries override to make it larger if a site actually needed it?

Thank you in advance,
Robin Holt
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/