PATCH: Diablo-style hashfeeds
"Miquel van Smoorenburg"
list-inn-workers at news.cistron.nl
Wed Jun 1 21:50:53 UTC 2005
I'm using INN to (help) feed a bunch of Diablo boxes. In the
diablo world, it is common to split one full feed over several
boxes using something called a "hashfeed". This patch adds
that functionality to innd.
( Russ, did you look at the innfeed-refcount-arthandle patch
I sent some time ago yet ? )
inn-2.4.2-hashfeed.patch
diff -ruN inn-2.4.2/doc/pod/newsfeeds.pod inn-2.4.2-new/doc/pod/newsfeeds.pod
--- inn-2.4.2/doc/pod/newsfeeds.pod 2004-12-22 05:21:19.000000000 +0100
+++ inn-2.4.2-new/doc/pod/newsfeeds.pod 2005-05-03 23:54:12.000000000 +0200
@@ -382,6 +382,29 @@
new process will run with. This flag can be used to raise the priority to
normal if you're using the I<nicekids> parameter in F<inn.conf>.
+=item B<Q> I<hashfeed>
+
+The hashfeed value for this site. The message-id of the article is
+hashed using a quick (and dirty) hash method (31-bits sum of the value
+of all characters). The hashfeed value must be given as value/mod
+or as start-end/mod. If the message-id hash modulus "mod" plus one
+equals "number" or is between "start" and "end", the article will be
+fed to the site.
+
+Example:
+
+ Q1/2 Feeds about 50% of the messages to this site
+ Q2/2 Feeds the other 50% of the messages
+
+ Q1-3/10 Feeds about 30% of the messages
+ Q4-5/10 Feeds about 30% of the messages
+ Q6/10 Feeds about 50% of the messages
+
+If this flag is specified multiple times the contents will be
+logically ORed together (just one match needed).
+
+The algorithm is compatible with the one used by Diablo.
+
=item B<S> I<size>
If the amount of data queued for the site gets to be larger than I<size>
diff -ruN inn-2.4.2/innd/art.c inn-2.4.2-new/innd/art.c
--- inn-2.4.2/innd/art.c 2004-12-22 05:21:19.000000000 +0100
+++ inn-2.4.2-new/innd/art.c 2005-05-03 23:36:05.000000000 +0200
@@ -1547,6 +1547,29 @@
}
/*
+** Return true if an element of the QHASHLIST matches the
+** quickhash of the message-id.
+*/
+static bool
+QHashMatch(QHASHLIST *qh, char *MessageID)
+{
+ unsigned char *p = (unsigned char *)MessageID;
+ int h = 0;
+
+ while (*p)
+ h += *p++;
+
+ while (qh) {
+ if ((h % qh->mod + 1) >= qh->begin &&
+ (h % qh->mod + 1) <= qh->end)
+ return true;
+ qh = qh->next;
+ }
+
+ return false;
+}
+
+/*
** Propagate an article to the sites have "expressed an interest."
*/
static void
@@ -1619,6 +1642,10 @@
* cross-posting. */
continue;
+ if (sp->QHashList && !QHashMatch(sp->QHashList, HDR(HDR__MESSAGE_ID)))
+ /* quickhash doesn't match */
+ continue;
+
if (list && *list != NULL && sp->Distributions &&
!DISTwantany(sp->Distributions, list))
/* Not in the site's desired list of distributions. */
diff -ruN inn-2.4.2/innd/innd.h inn-2.4.2-new/innd/innd.h
--- inn-2.4.2/innd/innd.h 2004-12-22 05:21:19.000000000 +0100
+++ inn-2.4.2-new/innd/innd.h 2005-05-03 23:36:29.000000000 +0200
@@ -407,6 +407,17 @@
/*
+** Diablo-style hashed feeds or hashfeeds.
+*/
+typedef struct _QHASHLIST {
+ int begin;
+ int end;
+ int mod;
+ struct _QHASHLIST *next;
+} QHASHLIST;
+
+
+/*
** A site may reject something in its subscription list if it has
** too many hops, or a bad distribution.
*/
@@ -458,6 +469,7 @@
struct buffer Buffer;
bool Buffered;
char ** Originator;
+ QHASHLIST * QHashList;
int Next;
int Prev;
} SITE;
diff -ruN inn-2.4.2/innd/newsfeeds.c inn-2.4.2-new/innd/newsfeeds.c
--- inn-2.4.2/innd/newsfeeds.c 2004-12-22 05:21:19.000000000 +0100
+++ inn-2.4.2-new/innd/newsfeeds.c 2005-05-04 00:01:57.000000000 +0200
@@ -448,6 +448,7 @@
int isp;
SITE *nsp;
struct buffer b;
+ QHASHLIST *qh;
b = sp->Buffer;
*sp = SITEnull;
@@ -467,6 +468,7 @@
sp->NeedOverviewCreation = false;
sp->FeedwithoutOriginator = false;
sp->DropFiltered = false;
+ sp->QHashList = NULL;
/* Nip off the first field, the site name. */
if ((f2 = strchr(Entry, NF_FIELD_SEP)) == NULL)
@@ -603,6 +605,18 @@
if (*++p && CTYPE(isdigit, *p))
sp->Nice = atoi(p);
break;
+ case 'Q':
+ qh = xmalloc(sizeof(QHASHLIST));
+ p++;
+ if (sscanf(p, "%d/%d", &qh->begin, &qh->mod) == 2) {
+ qh->end = qh->begin;
+ } else if (sscanf(p, "%d-%d/%d", &qh->begin, &qh->end, &qh->mod) != 3) {
+ free(qh);
+ return "hash not in x/z or x-y/z format";
+ }
+ qh->next = sp->QHashList;
+ sp->QHashList = qh;
+ break;
case 'S':
if (*++p && CTYPE(isdigit, *p))
sp->StartSpooling = atol(p);
diff -ruN inn-2.4.2/innd/site.c inn-2.4.2-new/innd/site.c
--- inn-2.4.2/innd/site.c 2004-12-22 05:21:19.000000000 +0100
+++ inn-2.4.2-new/innd/site.c 2005-06-01 23:41:42.000000000 +0200
@@ -997,6 +997,7 @@
SITEfree(SITE *sp)
{
SITE *s;
+ QHASHLIST *qh, *qn;
int new;
int i;
@@ -1051,6 +1052,13 @@
sp->FNLnames.data = NULL;
sp->FNLnames.size = 0;
}
+ if (sp->QHashList) {
+ for (qh = sp->QHashList; qh; qh = qn) {
+ qn = qh->next;
+ free(qh);
+ }
+ sp->QHashList = NULL;
+ }
/* If this site was a master, find a new one. */
if (sp->IsMaster) {
--
The From: and Reply-To: addresses are internal news2mail gateway addresses.
Reply to the list or to "Miquel van Smoorenburg" <miquels at cistron.nl>
More information about the inn-workers
mailing list