Load times for Large Subnets
Larry Apolonio
isc-dhcp at rh73.com
Tue Oct 9 21:22:49 UTC 2007
David W. Hankins wrote:
>On Tue, Oct 09, 2007 at 10:51:12AM -0700, Larry Apolonio wrote:
>
>
>>Basically I up'd that number by an order of magnitude + some (I think it
>>was 20000003) and my load times dropped from 20 minutes to 6 seconds.
>>
>>
>
>Are you sure you also aren't comparing across-feature-trains? E.g.,
>3.0.x vs 3.1.x?
>
>If not I wonder if we still have some lease hash collision problems,
>and you'd be doing me a favor if you could rebuild with
>REPORT_HASH_PERFORMANCE defined and the default hash table size (that
>only works in 3.1 or 4.0).
>
>Going from 3:1 to 1:48 buckets shouldn't net such wide results.
>
>
>
>>This section was in dhcp-3.1.0 and dhcp-4.0.0.a3 but it was not in the
>>RHEL SRPMS dhcp-3.0.5.
>>
>>
>
>3.0.x uses 'DEFAULT_HASH_SIZE' for everything, 9973. You could up it,
>but you'd up it for -all hash tables-.
>
>
>
>>There certainly was a huge cost in memory, I think it took nearly a GB
>>of RAM, but dhcperf showed similar results in 4 way handshakes.
>>
>>
>
>That hash size is used for several tables, and consumes a pointer
>for each bucket.
>
>So it's a rather multiplicative increase in memory footprint.
>
>
>
>>Does anyone know if there are any other problems that will arise when I
>>increased this number? What other tweaks are there that I can do to
>>improve transaction performance. Would be nice to be able to do 100,000
>>four way handshakes per hour.
>>
>>
>
>If you're going to run on tmpfs, and you're going to be editing
>sources anyway, comment out the fsync() call. It's an extra system
>call at that point. Not doing anyone any good.
>
>
>
Here is the recompiled results with report hash performance
Oct 9 14:20:47 dhcp-amd root: ==================================
Oct 9 14:20:50 dhcp-amd dhcpd: Internet Systems Consortium DHCP Server
V3.1.0
Oct 9 14:20:50 dhcp-amd dhcpd: Copyright 2004-2007 Internet Systems
Consortium.
Oct 9 14:20:50 dhcp-amd dhcpd: All rights reserved.
Oct 9 14:20:50 dhcp-amd dhcpd: For info, please visit
http://www.isc.org/sw/dhcp/
Oct 9 14:20:50 dhcp-amd dhcpd: DHCP name hash: Contents/Size (%):
95/401 (23%). Min/max: 0/2
Oct 9 14:20:50 dhcp-amd dhcpd: DHCP code hash: Contents/Size (%):
95/254 (37%). Min/max: 0/1
Oct 9 14:20:50 dhcp-amd dhcpd: NWIP name hash: Contents/Size (%): 11/17
(64%). Min/max: 0/2
Oct 9 14:20:50 dhcp-amd dhcpd: NWIP code hash: Contents/Size (%): 11/17
(64%). Min/max: 0/1
Oct 9 14:20:50 dhcp-amd dhcpd: FQDN name hash: Contents/Size (%): 8/13
(61%). Min/max: 0/2
Oct 9 14:20:50 dhcp-amd dhcpd: FQDN code hash: Contents/Size (%): 8/13
(61%). Min/max: 0/1
Oct 9 14:20:50 dhcp-amd dhcpd: VIVCO name hash: Contents/Size (%):
1/127 (0%). Min/max: 0/1
Oct 9 14:20:50 dhcp-amd dhcpd: VIVCO code hash: Contents/Size (%):
1/127 (0%). Min/max: 0/1
Oct 9 14:20:50 dhcp-amd dhcpd: VIVSO name hash: Contents/Size (%):
1/127 (0%). Min/max: 0/1
Oct 9 14:20:50 dhcp-amd dhcpd: VIVSO code hash: Contents/Size (%):
1/127 (0%). Min/max: 0/1
Oct 9 14:20:50 dhcp-amd dhcpd: ISC name hash: Contents/Size (%): 2/3
(66%). Min/max: 0/1
Oct 9 14:20:50 dhcp-amd dhcpd: ISC code hash: Contents/Size (%): 2/3
(66%). Min/max: 0/1
Oct 9 14:20:50 dhcp-amd dhcpd: Relay Agent name hash: Contents/Size
(%): 5/11 (45%). Min/max: 0/1
Oct 9 14:20:50 dhcp-amd dhcpd: Relay Agent code hash: Contents/Size
(%): 5/11 (45%). Min/max: 0/1
Oct 9 14:20:50 dhcp-amd dhcpd: Server-Config Option name hash:
Contents/Size (%): 52/106 (49%). Min/max: 0/3
Oct 9 14:20:50 dhcp-amd dhcpd: Server-Config Option code hash:
Contents/Size (%): 52/106 (49%). Min/max: 0/1
Oct 9 14:20:55 dhcp-amd dhcpd: Wrote 0 deleted host decls to leases file.
Oct 9 14:20:56 dhcp-amd dhcpd: Wrote 0 new dynamic host decls to leases
file.
Oct 9 14:20:57 dhcp-amd dhcpd: Wrote 200000 leases to leases file.
Oct 9 14:20:57 dhcp-amd dhcpd: Host HW hash: Contents/Size (%):
6/22501 (0%). Min/max: 0/2
Oct 9 14:20:57 dhcp-amd kernel: dhcpd[10365]: segfault at
0000000000000000 rip 0000000000481427 rsp 00007fff15f5c9d0 error 4
Oct 9 14:20:57 dhcp-amd dhcpd: dhcpd startup failed
More information about the dhcp-users
mailing list