ZTP - guesthell does not load after loading startup script

[_|brt.drml|_]

Hi, 

Would you like to enter the initial configuration dialog? [yes/no]:
Loading 9K/python_script.py from 10.15.255.52 (via Vlan990): !
[OK - 8646 bytes]

If I use the debug from my tftpd server I get this:

Connection received from 10.242.67.27 on port 52391 [14/10 13:03:34.105]
Read request for file <python_script.py>. Mode octet [14/10 13:03:34.106]
OACK: <blksize=8192,> [14/10 13:03:34.106]
Using local port 49808 [14/10 13:03:34.106]
Peer returns ERROR <Session terminated> -> aborting transfer [14/10 13:03:34.118]
Connection received from 10.242.67.27 on port 50484 [14/10 13:03:34.119]
Read request for file <python_script.py>. Mode octet [14/10 13:03:34.119]
OACK: <blksize=8192,> [14/10 13:03:34.119]
Using local port 49809 [14/10 13:03:34.119]
Connection received from 10.242.67.27 on port 54963 [14/10 13:03:36.119]
Read request for file <python_script.py>. Mode octet [14/10 13:03:36.120]
OACK: <blksize=8192,> [14/10 13:03:36.120]
Using local port 49810 [14/10 13:03:36.120]
Peer returns ERROR <Session terminated> -> aborting transfer [14/10 13:03:36.143]
Connection received from 10.242.67.27 on port 51977 [14/10 13:03:36.144]
Read request for file <python_script.py>. Mode octet [14/10 13:03:36.144]
OACK: <blksize=8192,> [14/10 13:03:36.145]
Using local port 49811 [14/10 13:03:36.145]
Connection received from 10.242.67.27 on port 53283 [14/10 13:03:38.152]
Read request for file <python_script.py>. Mode octet [14/10 13:03:38.152]
OACK: <blksize=8192,> [14/10 13:03:38.152]
Using local port 49812 [14/10 13:03:38.152]
Peer returns ERROR <Session terminated> -> aborting transfer [14/10 13:03:38.165]
Connection received from 10.242.67.27 on port 51878 [14/10 13:03:38.166]
Read request for file <python_script.py>. Mode octet [14/10 13:03:38.167]
OACK: <blksize=8192,> [14/10 13:03:38.167]
Using local port 49813 [14/10 13:03:38.167]
Connection received from 10.242.67.27 on port 52056 [14/10 13:03:40.168]
Read request for file <python_script.py>. Mode octet [14/10 13:03:40.168]
OACK: <blksize=8192,> [14/10 13:03:40.169]
Using local port 49814 [14/10 13:03:40.169]
Peer returns ERROR <Session terminated> -> aborting transfer [14/10 13:03:40.181]
Connection received from 10.242.67.27 on port 53005 [14/10 13:03:40.182]
Read request for file <python_script.py>. Mode octet [14/10 13:03:40.183]
OACK: <blksize=8192,> [14/10 13:03:40.183]
Using local port 49815 [14/10 13:03:40.183]
Connection received from 10.242.67.27 on port 54992 [14/10 13:03:42.182]
Read request for file <python_script.py>. Mode octet [14/10 13:03:42.183]
OACK: <blksize=8192,> [14/10 13:03:42.183]
Using local port 49816 [14/10 13:03:42.183]
Peer returns ERROR <Session terminated> -> aborting transfer [14/10 13:03:42.201]
Connection received from 10.242.67.27 on port 49485 [14/10 13:03:42.202]
Read request for file <python_script.py>. Mode octet [14/10 13:03:42.203]
OACK: <blksize=8192,> [14/10 13:03:42.203]
Using local port 49817 [14/10 13:03:42.203]
<python_script.py>: sent 2 blks, 8628 bytes in 0 s. 0 blk resent [14/10 13:03:42.229]
TIMEOUT waiting for Ack block #0  [14/10 13:03:49.124]
TIMEOUT waiting for Ack block #0  [14/10 13:03:51.149]
TIMEOUT waiting for Ack block #0  [14/10 13:03:53.169]
TIMEOUT waiting for Ack block #0  [14/10 13:03:55.185]

 The script is not downloaded, resulting in a fail and the switch starts in default mode.

Any idea? 

[_|brt.drml|_ · Accepted solution]

Solved it, 

The issue was that the tftp blocksize was defaulted to 8192.

On my network the packets got fragmented and therefore the script download failed. 

Reducing to a lower blocksize did not work either. Why: when rebooting, after wr erase the c9300 switch to its defaults, the blocksize was again set to default.

i enabled a webs server with http:80 and then the start-up did work again.

As a solution I enabled the start-up on the web-server and the script can then fetch some files on the tftp to complete the installation.