[UDEL Mail System: Failed mail (msg.aa25782)]


Mills@UDEL.EDU
Mon, 20 Jun 88 16:15:28 EDT


Ooop.

----- Forwarded message # 1:

Date: Mon, 20 Jun 88 14:33:53 EDT
From: UDEL Mail System (04-Jun-88/MMDF-2b_p#30) <mmdf@UDEL.EDU>
Sender: Mail System (MMDF) <mmdf@UDEL.EDU>
Subject: Failed mail (msg.aa25782)
To: Mills@UDEL.EDU
Message-ID: <8806201433.aa25894@Huey.UDEL.EDU>

    Your message could not be delivered to
'tcp-ip@sri.nic.arpa (host: Louie.UDEL.EDU) (queue: badhosts)' for the following
reason: ' (BHST) Unknown host/domain name in "tcp-ip@sri.nic.arpa"'

    Your message follows:

Date: Mon, 20 Jun 88 14:31:37 EDT
From: From: Mills@UDEL
To: tcp-ip@sri.nic.arpa, ineng-interest@isi.edu, ntp@trantor.umd.edu
Subject: NTP server status and peering strategies
Message-ID: <8806201431.aa25782@Huey.UDEL.EDU>

Folks,

First, let me apologize for the length of this message. I am sending it to the
entire list because it applies to a surprisingly large number of hosts
wittingly or unwittingly chiming the Network Time Protocol (NTP) and this may
be the only effective way to reach them. The six NTP primary time servers are
rapidly filling up with customers, so there is need for some discipline in
access. This message outlines corrective strategies and lists the current
servers, clients and points of contact. If these arcane issues do not tinkle
your bell, you might want to discard this message before reading. If so,
please be advised the ntp@trantor.umd.edu list is the official bazaar for NTP.

Dave

==============================================================================

For your information, at the end of this message are billboards showing the
NTP peers of various fuzzballs known to me. They show the size of the peer
population escalating sharply and cause some concern that server congestion
may begin to occur sooner than expected. Accordingly, I would like to reaffirm
the peerage suggestions made previously on this list and in the document
defining the new Version 1 of NTP (which was submitted for consideration as an
RFC, but has not yet seen the day of light). As a reminder, that document can
be found in pub/ntp.doc on louie.udel.edu.

First, all Unix chimers (are there any others?) should be using the latest
ntpd daemon from Mike Petry, which supports variable-rate polling and can
reduce the polling overhead by a factor of sixteen. Early versions of ntpd do
not provide the best filtering and deglitching features and do not
automatically adapt to failure modes.

Second, unless there are compelling reasons to the contrary, no more than two
secondary servers per network should rattle the five primary server cages and
all retail (end-user) hosts should obtain time from one or more secondary
servers. If two secondary servers are used, each of these should peer with two
different primary servers, for a total of four distinct primary servers. The
two secondary servers should peer with each other and with one or two
secondary servers in other networks. Experience suggests that the peering
population to stratum-one and stratum-two servers should be restrained;
otherwise, the clock-selection algorithm can allow a gang of secondary servers
to overwhelm a few primary servers.

Third, as shown in the following billboards, some chimers may be tinkling the
bells of inappropriate servers. The net 128.59 dudes tinkling udel2.udel.edu
(192.5.39.88) aka dcn6.arpa (128.4.0.6) are actually getting stratum-three
time or worse, although the time is currently pretty good. This host happens
to be located at my home and is vulnerable to new, untried software, power
glitches and marauding cats. Those dudes tinkling macom1.arpa (192.5.8.1,
10.0.0.111) may not be aware that the WWVB clock originally attached to that
host has been moved. However, while this host no longer provides stratum-one
time, it is appropriate for use as a stratum-two neighbor, as described above.
Also, note that the address for primary server ford1.arpa has changed from
128.5.0.1 to 128.5.192.1. I see a lot of ICMP error messages along the way to
128.5.0.1, suggesting that this fact is not well known.

Finally, note that primary server udel1.udel.edu (192.5.39.87) aka dcn1.arpa
(128.4.0.1, 10.2.0.96) is in process of being moved to a new address, which
should occur on or before 5 July. At the moment, the real WWVB clock is on
host udel7.udel.edu (192.5.39.94), so that udel1.udel.edu is currently
providing stratum-two time. A few of you guys cleverly figured that out and
switched chimes to the '94 address. Not to worry; all will be fixed soon. Be
advised that use of net number 192.5.39 is shortly to be discontinued along
with its ARPANET gateway. In order that the transition be as smooth as
possible, I suggest all chimers temporarily use ARPANET address 10.2.0.96.

If you are a secondary server chiming a primary server, you should find your
address in the following billboards. Hopefully, your net should appear no more
than once per billboard, assuming you conform to the peering suggestions
above. The format of these billboards are similar to those available with the
Unix daemon. Where points of contact are not listed, contact me for pointers.

==============================================================================

Host udel1.udel.edu is currently functioning as a secondary server
(stratum two). Its functions are being taken over by udel7.udel.edu, but all
addresses are shortly to be changed as mentioned above. Upon completion, this
host will become the recommended east-coast ARPANET primary server.

udel1.udel.edu (192.5.39.87, 10.2.0.96)
AdrID Address Port Status Delay Offset Disp Update
------------------------------------------------------------------------
0 [128.5.192.1] 123 377 756 64 95 16:48:58
1 [128.9.2.129] 123 377 732 133 297 16:49:12
2* [192.5.39.94] 123 377 35 -7 1 16:49:13
3 [128.153.4.2] 123 377 767 382 1269 16:48:52
4 [128.223.20.5] 123 000 659 -322 65535 01:10:36
5 [128.200.9.18] 123 377 451 35 160 16:48:55
6 [35.1.128.16] 123 377 419 -19 26 16:48:48
7 [26.6.0.2] 123 377 2135 -850 529 16:48:58
8 [128.45.1.1] 123 370 504 -6533 542 16:45:44
9 [128.8.251.56] 123 377 222 -317 64 16:48:58
10 [128.8.251.57] 123 377 256 -272 390 16:48:59
11 [26.18.0.124] 123 376 1600 -2808 791 16:48:49
12 [128.8.251.65] 123 377 208 -186 227 16:49:02
13 [128.8.251.64] 123 377 193 90 72 16:49:03
14 [10.1.0.17] 123 377 235 -1051 45 16:49:04
15 [128.200.16.5] 123 377 483 -190 62 16:49:05
16 [192.5.220.220] 123 377 211 415 101 16:49:06
17 [35.1.128.32] 123 376 346 -7 150 16:48:36
18 [192.5.57.1] 123 000 264 -2066 9 16:44:21
19 [128.200.16.7] 123 377 483 61 169 16:48:47
20 [15.255.64.2] 123 377 991 25 77 16:49:09
21 [128.8.251.67] 123 377 228 -278 338 16:48:45
22 [128.8.251.59] 123 377 228 -492 63 16:49:09
23 [128.8.10.14] 123 377 218 -6479 5553 16:49:10
24 [128.200.16.2] 123 377 517 -315 109 16:49:12
25 [128.2.249.123] 123 112 232 11 24 16:48:11
26 [128.8.251.58] 123 377 224 -474 123 16:49:14
27 [192.5.39.3] 123 377 23 63 44 16:49:15
28 [128.95.16.21] 123 377 490 -276 73 16:49:22
29 [128.175.2.13] 123 377 2 76 95 16:49:17
30 [128.8.251.61] 123 377 222 1703195130 53 16:49:18
31 [128.59.64.60] 123 003 319 373 32393 16:48:42
32 [129.48.1.1] 123 376 6669 -522 833 16:49:20
33 [128.174.5.50] 123 377 355 303 128 16:48:30
34 [10.0.0.31] 123 377 101 13 7 16:49:24
35 [128.8.251.51] 123 377 224 71 82 16:49:16
36 [128.208.1.14] 123 000 493 -35 65535 01:10:01
37 [128.116.64.3] 123 377 460 25 52 16:48:20
38 [128.8.251.60] 123 377 219 -100 208 16:49:13
39 [128.8.251.66] 123 377 238 -63 103 16:49:14
40 [128.49.16.7] 123 377 5321 -473 645 16:49:06
41 [128.200.16.6] 123 377 508 -125 101 16:48:29
42 [128.109.130.5] 3965 000 1275 -638 65022 08:05:05
43 [128.112.14.5] 123 377 95 3723 6 16:49:19
44 [128.112.14.25] 123 376 113 3741 6 16:48:27
45 [128.112.14.18] 123 000 102 -44954945 7180 14:01:21
46 [192.5.8.1] 123 377 166 -130 39 16:48:52
47 [15.255.17.2] 123 177 901 11 12 16:48:41
48 [128.59.64.60] 123 376 222 325 64 16:48:31
49 [128.95.16.22] 123 377 475 -340 250 16:48:41

Host wwvb.isi.edu is the recommended west-coast ARPANET/MILNET primary server.
Contact Steve Casner (casner@isi.edu) for further information.

wwvb.isi.edu (128.9.2.129)
AdrID Address Port Status Delay Offset Disp Update
------------------------------------------------------------------------
0 [128.116.64.3] 123 375 1204 64 442 17:39:52
1 [128.8.10.1] 123 377 1193 -18 577 17:34:40
2 [128.59.16.119] 123 376 1019 25 117 17:40:06
3 [128.200.16.7] 123 376 239 28 86 17:39:32
4* [192.5.39.94] 123 376 786 -149 422 17:40:06
5 [129.48.1.1] 123 376 5478 -501 911 17:39:59
6 [192.5.39.87] 123 376 710 -119 182 17:39:41
7 [128.200.16.2] 123 376 250 134 110 17:38:35
8 [128.59.16.125] 123 377 1029 -7 533 17:40:27
9 [128.59.16.111] 123 377 1092 156 474 17:40:30
10 [128.59.16.116] 123 377 1032 10 264 17:40:33
11 [128.59.16.112] 123 376 967 -6 249 17:40:12
12 [128.200.9.18] 123 216 228 25 41 17:39:49
13 [128.59.16.122] 123 376 1011 -156 183 17:39:50
14 [192.5.220.220] 123 377 894 -58 807 17:40:11
15 [128.200.16.5] 123 361 239 -64 149 17:38:39
16 [128.97.28.251] 123 376 141 -85 174 17:39:34
17 [128.59.16.120] 123 376 1020 -9 171 17:39:55
18 [192.12.252.4] 123 000 0 0 65535 17:39:43
19 [15.255.64.2] 123 367 1772 100 136 17:40:24
20 [128.200.16.6] 123 376 251 -41 146 17:40:19
21 [128.5.192.1] 123 376 2328 -98 1255 17:24:26
22 [15.255.17.2] 123 311 1583 17 225 17:39:05

Host ncar.nsf.net is one of the NSFNET Backbone switches. As most already
know, the NSFNET Backbone is shortly to be retired and replaced by the
Merit/IBM/MCI NSFNET Backbone, so that the ncar.nsf.net fuzzy will no longer
have that mission. Hopefully, this host will remain in place, if only to
provide good time and remain the recommended NSFNET midcontinent primary
time server.

ncar.nsf.net (128.116.64.3)
AdrID Address Port Status Delay Offset Disp Update
------------------------------------------------------------------------
0 [128.8.10.1] 123 376 397 -99 1 04:58:15
1 [128.4.0.1] 123 377 422 -16 132 05:09:21
2 [128.135.4.2] 123 376 136 -3 22 05:08:49
3* [192.5.39.94] 123 376 403 -13 99 05:09:05
4 [192.5.220.220] 123 305 550 -226 216 05:07:56
5 [128.200.16.7] 123 376 145 103 81 05:09:00
6 [128.135.20.72] 123 376 133 -19 9 05:08:57
7 [128.200.16.6] 123 376 153 -400 217 05:08:16
8 [128.138.240.4] 123 377 59 0 8 05:09:01
9 [15.255.64.2] 123 376 1168 29 72 05:09:34
10 [128.200.16.3] 123 377 158 55 297 05:09:04
11 [128.174.5.50] 123 377 93 1 7 05:09:30
12 [128.223.20.5] 123 337 418 32 142 05:09:39
13 [128.102.18.20] 123 173 206 308 193 05:09:32
14 [128.200.9.18] 123 376 159 -222 110 05:09:09
15 [128.200.16.5] 123 000 165 -90 65535 12:37:04
16 [128.200.16.2] 123 377 141 28 55 05:09:33
17 [128.42.1.64] 123 377 79 -2 4 05:09:08
18 [128.9.2.129] 123 136 1532 -17 843 04:53:47
19 [15.255.16.12] 123 327 1078 20 73 05:03:33
20 [15.255.17.2] 123 361 1038 23 47 05:08:14
21 [128.5.192.1] 123 376 900 82 116 05:01:48
22 [15.255.64.2] 2609 000 836 -418 65022 23:33:11

Ramshackle host ford1.arpa is presently at the end of a 9600-bps line at
Dearborn, MI. Net 128.5 should soon be rehomed to MERIT (net 35) by a 56-Kbps
line, in which case (net managers willing) it should become a recommended
NSFNET midwestern primary server. It would be most polite if potential
chimers were to contact Fred Ball (ball@ford-vax.arpa) before use.

ford1.arpa (128.5.192.1)
AdrID Address Port Status Delay Offset Disp Update
------------------------------------------------------------------------
0 [128.9.2.129] 123 376 2208 -450 197 19:14:13
1 [128.116.64.3] 123 127 1160 92 173 19:26:40
2 [192.5.39.94] 123 333 667 -117 26 19:26:01
3 [128.59.16.121] 123 376 781 89 435 19:25:58
4 [192.5.220.220] 123 126 984 -67 198 19:26:02
5 [192.5.39.87] 123 376 762 -99 93 19:25:39
6 [15.255.17.2] 123 376 1635 -37 210 19:23:59
7 [15.255.16.12] 123 375 1575 -115 113 19:25:54
8 [128.59.16.1] 123 377 971 79 436 19:26:20
9* [192.5.8.1] 123 367 269 -36 46 19:26:07
10 [128.59.16.116] 123 377 794 50 217 19:26:34
11 [128.59.16.119] 123 376 935 12 323 19:26:00
12 [128.143.2.7] 123 376 807 -21 175 19:26:03
13 [128.59.16.122] 123 376 630 -30 157 19:26:25
14 [128.59.16.112] 123 377 806 -39 123 19:26:42
15 [128.59.16.125] 123 367 688 -19 209 19:26:45
16 [128.59.16.120] 123 375 763 -79 73 19:25:56
17 [128.59.16.111] 123 377 891 -186 324 19:25:58
18 [128.97.28.251] 123 361 1253 -153 302 19:25:04
19 [128.120.57.23] 2560 000 0 0 65535 19:15:07

Host umd1.umd.edu has become the busiest primary server; however, access via
ARPANET is shortly to be curtailed. This is the recommended east-coast NSFNET
MILNET primary server. Note that this server has 63 clients, which is just two
shy of the 65-client maximum of the present configuration. Contact Mike Petry
(petry@trantor.umd.edu) or Louie Mamakos (louie@trantor.umd.edu) for further
information.

umd1.umd.edu (128.8.10.1)
AdrID Address Port Status Delay Offset Disp Update
------------------------------------------------------------------------
0 [128.5.192.1] 123 346 627 45 148 17:46:49
1 [128.9.2.129] 123 377 992 38 156 17:47:22
2 [128.116.64.3] 123 052 581 71 26 17:39:58
3 [128.59.16.111] 123 376 232 -50 53 17:49:29
4 [128.8.251.57] 123 377 18 -347 112 17:49:15
5 [128.59.16.116] 123 377 189 -38 86 17:49:33
6 [128.143.2.7] 123 377 71 0 4 17:49:33
7 [128.200.16.2] 123 212 636 -34 479 17:40:17
8 [128.95.16.22] 123 377 1011 5 44 17:49:32
9 [35.1.128.32] 123 367 82 5 8 17:49:26
10 [128.8.251.51] 123 377 13 44 149 17:48:48
11 [128.59.64.60] 123 000 160 275 65535 16:47:45
12 [128.45.1.1] 123 000 570 -6419 65535 16:45:09
13 [35.1.128.16] 123 377 76 43 14 17:49:35
14 [128.59.16.122] 123 377 187 -56 87 17:48:50
15 [128.8.251.61] 123 377 14 1703195154 13 17:49:36
16 [36.22.0.8] 123 376 643 29 162 17:48:44
17 [35.1.128.16] 2289 000 0 0 65535 05:00:58
18 [128.59.16.120] 123 377 188 -37 51 17:48:55
19 [128.8.1.14] 123 376 21 -1174 14 17:48:35
20* [192.5.39.94] 123 377 173 -47 37 17:49:01
21 [10.0.0.31] 123 376 241 -2 33 17:48:39
22 [128.223.20.5] 123 000 1129 -266 65535 01:10:55
23 [128.200.9.18] 123 342 545 -143 175 17:46:40
24 [192.5.39.88] 123 376 389 -38 14 17:49:22
25 [128.8.251.65] 123 376 24 -224 78 17:48:42
26 [129.48.1.1] 123 201 6141 -1077 1814 17:48:45
27 [128.8.251.64] 123 372 13 -308 198 17:48:42
28 [128.8.251.58] 123 376 25 -514 158 17:48:41
29 [128.8.251.60] 123 376 8 -407 218 17:48:43
30 [128.8.251.56] 123 376 19 -246 123 17:48:45
31 [128.59.16.119] 123 376 191 -14 46 17:49:05
32 [128.95.16.21] 123 376 752 51 286 17:48:50
33 [128.97.28.251] 123 377 616 -425 733 17:49:40
34 [128.135.20.72] 123 377 508 -16 65 17:49:11
35 [129.2.1.6] 123 377 28 1 0 17:49:34
36 [128.8.132.10] 123 372 23 -2 2 17:49:13
37 [18.71.0.8] 123 376 0 0 65535 17:48:54
38 [128.175.2.13] 123 376 50 186 156 17:48:55
39 [128.200.16.5] 123 376 463 177 161 17:48:59
40 [128.8.251.66] 123 376 15 -463 110 17:48:55
41 [128.135.4.2] 123 376 348 -40 39 17:49:17
42 [128.8.251.67] 123 376 27 -347 328 17:48:59
43 [192.5.39.3] 123 376 143 37 17 17:48:58
44 [128.49.16.7] 123 377 3141 -2603 3143 17:49:31
45 [128.102.22.200] 123 377 918 -591 249 17:49:00
46 [128.8.251.59] 123 377 25 -439 236 17:49:06
47 [128.8.10.14] 123 377 16 -45 335 17:49:15
48 [128.59.16.125] 123 377 207 -56 44 17:49:27
49 [128.200.16.7] 123 375 501 49 279 17:49:17
50 [18.26.0.109] 123 377 0 0 65535 17:49:09
51 [128.153.4.2] 123 377 1130 -85 585 17:49:09
52 [192.5.220.220] 123 377 236 -67 2271 17:49:34
53 [128.102.18.20] 123 351 929 93 491 17:44:23
54 [128.42.1.64] 123 365 280 -5 17 17:44:46
55 [128.138.240.4] 123 154 387 -60 262 17:41:10
56 [128.200.16.6] 123 376 529 -269 350 17:49:09
57 [128.112.14.5] 123 377 208 -12 16 17:49:42
58 [128.112.14.25] 123 377 204 3696 15 17:45:12
59 [128.112.14.18] 123 000 0 0 65535 14:01:21
60 [192.5.8.1] 123 163 236 -103 26 17:49:34
61 [15.255.17.2] 123 376 1031 45 55 17:48:49
62 [128.95.16.22] 123 000 762 122 65535 17:35:27
63 [128.59.16.112] 123 376 237 -6 75 17:49:11

As mentioned above, host udel1.udel.edu is not presently a primary server, but
obtains primary time from the following host udel7.udel.edu. On or before 5
July, this the address of udel7.udel.edu will be changed and it will become
the official primary server instead of udel1.udel.edu. While the existence of
udel7.udel.edu has not been officially revealed, some clever sods found it
anyway, probably by chasing the IP address field of NTP updates.

udel7.udel.edu (192.5.39.94)
AdrID Address Port Status Delay Offset Disp Update
------------------------------------------------------------------------
0 [128.5.192.1] 123 375 758 112 194 15:21:30
1 [128.9.2.129] 123 376 726 142 49 15:21:42
2 [128.8.10.1] 123 366 180 51 17 15:21:43
3 [128.116.64.3] 123 267 518 14 92 15:21:51
4* [18.72.0.3] 123 377 190 -25 15 15:22:12
5 [128.8.10.14] 123 377 158 81 555 15:21:45
6 [18.26.0.109] 123 364 212 -21 20 15:07:18
7 [192.5.39.87] 123 376 35 6 0 15:21:37
8 [18.71.0.8] 123 365 197 -81 110 15:15:34
9 [35.1.128.32] 123 215 296 98 12 15:20:41
10 [18.72.0.141] 3106 000 14633 -7317 65022 16:35:39
11 [35.1.128.16] 123 326 218 59 149 15:07:45
12 [18.72.0.106] 2333 000 181 -91 15967 00:53:31

As noted, host macom1.arpa is no longer a primary server. I assume the
customers listed know that. Contact Mike Little (little@macom2.arpa) for
further information.

macom1.arpa (192.5.8.1, 10.0.0.111)
AdrID Address Port Status Delay Offset Disp Update
------------------------------------------------------------------------
0 [128.5.192.1] 123 375 294 171 43 13:23:45
1* [128.8.10.1] 123 377 226 169 19 13:24:22
2 [128.4.0.1] 123 376 152 161 22 13:24:23
3 [10.0.0.31] 123 125 224 125 117 13:09:54
4 [18.62.0.7] 123 125 276 120 255 13:24:36
5 [15.255.17.2] 123 053 1058 205 391 13:17:44
6 [15.255.16.12] 123 135 1030 150 13 13:24:48
7 [18.62.0.6] 123 000 182 -1030632 65535 08:20:52
8 [128.109.131.49] 1198 000 257 -129 24700 03:55:10
9 [128.109.130.5] 4126 000 181 -91 9859 08:02:11

Host udel2.udel.edu is probabably not an appropriate secondary server, so its
net 128.59 customers should move elsewhere. Also, these customers should also
upgrade to the latest ntpd. The other customers are mine.

udel2.udel.edu (192.5.39.88) aka dcn6.arpa (128.4.0.6)
AdrID Address Port Status Delay Offset Disp Update
------------------------------------------------------------------------
0 [128.8.10.1] 123 373 383 61 13 01:49:39
1 [128.59.16.116] 123 376 469 -29 32 01:50:33
2 [128.59.16.112] 123 377 467 -9 13 01:50:39
3 [128.59.16.122] 123 377 465 -8 15 01:49:50
4 [128.59.16.120] 123 377 480 -10 32 01:49:55
5 [128.59.16.119] 123 377 491 1 46 01:50:05
6* [128.59.16.125] 123 377 455 -8 15 01:50:26
7 [128.59.16.111] 123 377 479 -13 43 01:50:30
8 [128.175.2.13] 123 377 231 132 61 01:50:00

==============================================================================

----- End of forwarded messages



This archive was generated by hypermail 2.0b3 on Thu Mar 09 2000 - 14:42:31 GMT