Back Ground:
I faced the issue that module have never been returned after “Testing connection” message by sending “cm data connection”.
[2020-07-02 08:59:17.431] root@swi-mdm9x28-wp:~# cm data connect
[2020-07-02 08:59:24.531] Setting up profile 1
[2020-07-02 08:59:24.548] Connecting … ok
[2020-07-02 08:59:24.707] Checking if device ‘rmnet_data0’ is up … ok
[2020-07-02 08:59:24.803] Routing inet … ok
[2020-07-02 08:59:24.820] mv: can’t remove ‘/etc/resolv.conf’: Device or resource busy
[2020-07-02 08:59:24.841] Updating /etc/resolv.conf … ok
[2020-07-02 08:59:24.841] Updating /etc/resolv.conf … ok
[2020-07-02 08:59:24.841] Testing connection … <== stopped
So I set [ctrl+C] in the timing of "Testing connection".
However, module can establish PDP context after that, and succeeded ping.
I keep to wait any returns after that, so module was returned "failed" after 10mins from "Testing connection" message.
However, module can also establish PDP context after that.
====
I found potential work around that script of “cm” was changed source code(legato/systems/current/apps/tools/read-only/scripts/cm) as follows.
comment out Line#212 and #213
echo -n “Testing connection …”
curl --silent $TEST_SITE > $LOG 2>&1
As the result, module can be returned “ok” after “Testing connection” quickly.
Q1:
Is there any influence to be comment out as above?
Q2:
Do you know the know behavior of this question?
As jyijyi indicated, if you can ping after the connect then all is well.
cm tries to connect to https://www.akamari.com which is slow and bulky and in my experience often down. There also seem to be issue with https and curl (not sure yet… seems ipv6 related)
You can change TEST_SITE to http://www.google.com in the cm script to get quick pass result if you want.
Thanks for reply.
I understand that module can be succeeded ping.
However, I could not notify the success of cm data connect, so I could not proceed next sequence.
I don’t focus on NW connection issue.
As jyijyi said, I understand cm command could not use for end production, so I would like to know the influence in the fixing source code point of view.
OK, I understand there is no influence by adding comment out as follows.
legato/systems/current/apps/tools/read-only/scripts/cm
// echo -n “Testing connection …”
// curl --silent $TEST_SITE > $LOG 2>&1
comment out Line#212 and #213
echo -n “Testing connection …”
curl --silent $TEST_SITE > $LOG 2>&1
BTW, could you tell me more detail what process does the above operate?
Currently, we try to implement LE function, so we checked source code.
If you know, could you please let me know it?
Sorry for rushing you.
Yes, I understanding the summarize, but I would like to know your plan whether you will fix the behavior of test connection to akamari.com in future legato version.
Of course, I understand that CM command is used for test purpose, but I concerned whether is any influence due to remaining the behavior of test connection to akamari.com.
(Other Sierra engineer said it seems to be caused MCU busy by analyzing logs, and this “No return” behavior was caused in all of WP7605.
That’s why, I asked it. Sorry for inconvenient you.