Request for more logging on node reporting

[skyler440]

I am dealing with identifying a potential bug, where the list of non-adjacent nodes gets "stuck" during a disconnect, and prevents us from being able to bring the connection back, which is negatively impacting our large ASL3 hub requiring frequent asterisk restarts to correct.

Even with the maximum debug, the only form of logging we get is the "playing remote-already.gsm" audio file. There is nothing logged about why it thinks this node is still connected, or which adjacent neighbor it might think its coming from.

This request is to increase the logging, in particular on this neighboring node list which is exchanged between different connected allstarlink nodes. If we had a way to report why the hub thinks these phantom nodes are still connected even though they are not, it would be very insightful.

[Allan-N]

https://community.allstarlink.org/t/asl3-loopback-protection-issues/24441

[Allan-N]

FYI: I "think" the code we need to review/rework is

app_rpt/apps/app_rpt/rpt_link.c

Lines 737 to 765 in a5c7a76

	} else { /* Check to see if this node is already linked */
	lstr = ast_str_create(RPT_AST_STR_INIT_SIZE);
	if (!lstr) {
	rpt_mutex_unlock(&myrpt->lock);
	goto cleanup;
	}
	n = __mklinklist(myrpt, NULL, &lstr, USE_FORMAT_RPT_LINK) + 1;
	rpt_mutex_unlock(&myrpt->lock);
	strs = ast_malloc(n * sizeof(char *));
	if (!strs) {
	ast_free(lstr);
	goto cleanup;
	}
	ns = finddelim(ast_str_buffer(lstr), strs, n);
	for (i = 0; i < ns; i++) {
	if ((*strs[i] < '0') || (*strs[i] > '9')) {
	strs[i]++;
	}
	if (!strcmp(strs[i], node)) {
	ast_free(lstr);
	ast_free(strs);
	rpt_telem_select(myrpt, connect_data->command_source, connect_data->mylink);
	rpt_telemetry(myrpt, REMALREADY, NULL);
	goto cleanup; /* Already linked */
	}
	}
	ast_free(strs);
	ast_free(lstr);
	}

[mkmer]

I think the "problem" is the non adjacent node is still in some remote systems list of connected nodes (Which may be ASL2 or NON ASL software) -> I don't think we can do much with it "locally" when/if that happens.

[Allan-N]

I don't see this PR as fixing the problem (if one actually exists). This is more about more clearly explaining why we are saying "No" (e.g. we can't connect to node X because connected node Y is already linked to the node)