15 Apr
2016
15 Apr
'16
12:31 p.m.
I thought on Kamailio docs. Thanks for links.
So, workaround in my setup is to always call
sip_capture("$var(table_format)") with table format as argument, even don't
need to shard data, to get correct table name.
Kind regards and thanks,
Seudin
On Fri, Apr 15, 2016 at 10:03 AM, Alexandr Dubovikov <
notifications(a)github.com> wrote:
the docs are here:
https://github.com/sipcapture/homer/wiki
since H5 we generate table name depends on message type and datestamp:
I have pointed you already to the new kamailio's config:
https://github.com/sipcapture/homer-api/blob/master/examples/sipcapture/sip…
https://github.com/sipcapture/homer-api/blob/master/examples/sipcapture/sip…
https://github.com/sipcapture/homer-api/blob/master/examples/sipcapture/sip…
at the end (L855) you push the generated table name to sip_capture function
and a message will be inserted to the table.
modparam "table_name" is only for H3
Wbr,
Alexandr
On 15 April 2016 at 09:52, Seudin Kasumovic <notifications(a)github.com>
wrote:
--
MSC Seudin Kasumovic
Tuzla, Bosnia
---
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/kamailio/kamailio/issues/566#issuecomment-210382956
Hi,
I'm testing on 4.4.
The 'new design' isn't documented, so I can't figure out how the name
will
be generated.
When use simple config:
modparam("sipcapture", "db_url", DBURL)
modparam("sipcapture", "capture_on", 1)
modparam("sipcapture", "hep_capture_on", 1)
modparam("sipcapture", "raw_ipip_capture_on", 0)
modparam("sipcapture", "table_name", "sip_capture")
sipcapture uses given table name as format string, but the final result
was
'sip_capture.....A'. So no any timestamp
in the name. The reason for this
could be next two lines:
ntab.len = strftime(strftime_buf, sizeof(strftime_buf), table->s,
&capt_ts);
ntab.s = strftime_buf;
The table->s not necessary ends with '\0', because it's part of Kamailio
internal str structure.
Avoid non zero terminated string will work in this case.
Parsing loop in parse_table_names function doesn't add '\0' at the end of
table name.
Suggested patch is here.
diff --git a/modules/sipcapture/sipcapture.c
b/modules/sipcapture/sipcapture.c
index 41d153e..2bdb4b4 100644
--- a/modules/sipcapture/sipcapture.c
+++ b/modules/sipcapture/sipcapture.c
@@ -436,8 +436,9 @@ int parse_table_names (str table_name, str **
table_names){
{
LM_INFO ("INFO: table name:%s\n",p);
names[i].len = strlen (p);
- names[i].s = (char *)pkg_malloc(sizeof(char) *names[i].len);
+ names[i].s = (char *)pkg_malloc(sizeof(char) *(names[i].len + 1));
memcpy(names[i].s, p, names[i].len);
+ names[i].s[names[i].len] = '\0';
i++;
p = strtok (NULL, "| \t");
}
BTW, can you explain what 'table->s' will contain (format string required
by strftime) using provided example config ?
Thanks,
Seudin
—
You are receiving this because you commented.
Reply to this email directly or view it on GitHub
<https://github.com/kamailio/kamailio/issues/566#issuecomment-210343805>
—
You are receiving this because you authored the thread.
Reply to this email directly or view it on GitHub
<https://github.com/kamailio/kamailio/issues/566#issuecomment-210346518>