Issue

• NTP Server cannot synchronize with the upstream server if both have reference to its LOCAL clock
• NTP Server1 refers to its LOCAL clock(127.127.1.0) and refers to NTP Server2 and NTP Server2 refers to its LOCAL clock(127.127.1.0).

                 refer to LOCAL                  refer to LOCAL
                      ----                            ----
                     |    |                          |    |
                     V    |                          V    |
     ------------------   |       ---------------------   |
   | NTP Server1       |--       | NTP Server2         |  |
   | aqtta102          |-------->| aqttd102            |--
   | <IP Address 1>    |  refer  | <IP Address 2)     |
     ------------------           ---------------------

• The problem is that NTP Server1 could not synchronize with NTP Server2 but synchronized with its LOCAL clock. And we found that the error code "flash 0x0800" was reported in the above environment as follows:

 #ntpdc -c
 ntpdc>showpeer 10.248.250.62

 <snip>
 reach 377, unreach 0, flash 0x0800, boffset 0.00400, ttl/mode 0 
 <snip>                ~~~~~~~~~~~~~

• The following is the definition of the error code "0x0800".

 include/ntp.h
 -----------------------------------------------------------------------
 236 #define TEST12          0x0800  /* peer synchronization loop */
 -----------------------------------------------------------------------

• In the environment, other NTP clients, which refer to NTP Server1 and NTP Server2, could synchronize with NTP Server2. We think that NTP Server1 should not report the error code and should synchronize with NTP Server2.
• Is this a bug and what is the root cause?
• Is there workaround to not show the error?