recomputing skew time when receiving an advertisement from the master in rfc 5798
Vladica Stanisic <[email protected]> Mon, 17 Jan 2011 12:03:37 -0500
| Newsgroups | gmane.ietf.vrrp |
|---|---|
| Message-ID | <FFA917582FF6524F94AB504C6DC3F16DBD00BC204B@EUSAACMS0702.eamcs.ericsson.se> |
--===============1717328492== Content-Language: en-US Content-Type: multipart/alternative; boundary="_000_FFA917582FF6524F94AB504C6DC3F16DBD00BC204BEUSAACMS0702e_" --_000_FFA917582FF6524F94AB504C6DC3F16DBD00BC204BEUSAACMS0702e_ Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Based on the state machine in rfc 5798 on page 23 steps 450-460 when vrrp i= s in backup state and receives an advertisement there is no step to recompu= te the skew time. While in steps 740-765 on page 25 when vrrp is a master a= nd receives an advertisement from some new master, it does recompute the sk= ew time in step 750. Since skew time is part of master down interval, shouldn't it be recomputed= in both states when advertisement is received from a master? Otherwise bac= kup vrrp would continue to use skew time that was computed based on its co= nfigured value for advertisement interval when initially router transitions= to backup. Is this skew time recalculation implied in the backup state or there should= be an explicit step like in the master case? Vladica Stanisic --_000_FFA917582FF6524F94AB504C6DC3F16DBD00BC204BEUSAACMS0702e_ Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <HTML><HEAD> <META http-equiv=3DContent-Type content=3D"text/html; charset=3Dus-ascii"> <META content=3D"MSHTML 6.00.6001.18542" name=3DGENERATOR></HEAD> <BODY> <DIV> <DIV><SPAN class=3D248162321-11012011><FONT face=3DArial size=3D2>Based on&= nbsp;the=20 state machine in rfc 5798 on page 23 steps 450-460 when vrrp= is=20 in backup state and receives an advertisement there is no step to recompute= the=20 skew time. While in steps 740-765 on page 25 when vrrp is a maste= r and=20 receives an advertisement from some new master, it does recompute= the=20 skew time in step 750.</FONT></SPAN></DIV> <DIV><SPAN class=3D248162321-11012011></SPAN><SPAN class=3D248162321-110120= 11><FONT=20 size=3D2><FONT face=3DArial>Since skew time is part of master down=20 interval, shouldn't it be recomputed in both states when advertis= ement=20 is received from a master? Otherwise backup vrrp would continue to use skew= =20 time <SPAN class=3D774115216-17012011> </SPAN>that was computed based = on=20 its<SPAN class=3D774115216-17012011> </SPAN><SPAN=20 class=3D774115216-17012011>configured</SPAN> value for advertisement=20 interval<SPAN class=3D774115216-17012011> when initially router transitions= to=20 backup.</SPAN></FONT></FONT></SPAN></DIV> <DIV><SPAN class=3D248162321-11012011><FONT face=3DArial size=3D2><SPAN=20 class=3D774115216-17012011>Is this skew time recalculation implied in the b= ackup=20 state or there should be an explicit step like in the master=20 case?</SPAN></FONT></SPAN></DIV> <DIV><SPAN class=3D248162321-11012011><FONT face=3DArial=20 size=3D2></FONT></SPAN> </DIV> <DIV><SPAN class=3D248162321-11012011><SPAN class=3D774115216-17012011><FON= T=20 face=3DArial size=3D2>Vladica=20 Stanisic</FONT></SPAN></SPAN></DIV></DIV></BODY></HTML> --_000_FFA917582FF6524F94AB504C6DC3F16DBD00BC204BEUSAACMS0702e_-- --===============1717328492== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ vrrp mailing list [email protected] https://www.ietf.org/mailman/listinfo/vrrp --===============1717328492==--