[Fwd: clientside_tos and qos_flows]

From: Henrik Nordström <henrik_at_henriknordstrom.net>
Date: Tue, 27 Apr 2010 22:20:07 +0200

-------- Vidarebefordrat meddelande --------
Från: Jan Rheinländer <jrheinlaender_at_gmx.de>
Reply-to: resap-bdk_at_iamafg.org
Till: squid-bugs_at_squid-cache.org
Ämne: clientside_tos and qos_flows
Datum: Tue, 27 Apr 2010 10:23:46 +0430

Hi,

I tried the following:

clientside_tos 0x30 update
qos_flows local_hit=0x20 disable_preserve_miss

But no tos values 0x30 ever survived the qos_flows setting. So I suggest
changing client-side_reply.cc
(comm_set_tos(fd.tos); moved into the if clauses):

#if USE_ZPH_QOS
    if (reqofs==0 && !logTypeIsATcpHit(http->logType)) {
        assert(fd >= 0); // the beginning of this method implies fd may
be -1
        int tos = 0;
        if (Config.zph.tos_sibling_hit &&
http->request->hier.code==SIBLING_HIT ) {
            tos = Config.zph.tos_sibling_hit;
            debugs(33, 2, "ZPH: Sibling Peer hit with hier.code=" <<
http->request->hier.code << ", TOS=" << tos);
            comm_set_tos(fd,tos);
        } else if (Config.zph.tos_parent_hit &&
http->request->hier.code==PARENT_HIT) {
            tos = Config.zph.tos_parent_hit;
            debugs(33, 2, "ZPH: Parent Peer hit with hier.code=" <<
http->request->hier.code << ", TOS=" << tos);
            comm_set_tos(fd,tos);
        } else if (Config.zph.preserve_miss_tos &&
Config.zph.preserve_miss_tos_mask) {
            tos = fd_table[fd].upstreamTOS &
Config.zph.preserve_miss_tos_mask;
            debugs(33, 2, "ZPH: Preserving TOS on miss, TOS="<<tos);
            comm_set_tos(fd,tos);
        }
        //This breaks clientside_tos by ALWAYS setting tos to zero if
none of the above applies
        //Moved it into the if clauses
        //comm_set_tos(fd,tos);
    }
#endif

How requests which match acl update get 0x20, and local_hits get 0x30.

Sorry, the code is messy but the list won't accept HTML formatting.

Jan Rheinländer
Received on Tue Apr 27 2010 - 20:20:11 MDT

This archive was generated by hypermail 2.2.0 : Wed Apr 28 2010 - 12:00:48 MDT