andrewgroup Posted June 26, 2009 Report Share Posted June 26, 2009 With a reasonable amount of certainty we believe the CS410 will consistently crash when it sees 802.1D spanning tree packets. The CS410 creates the following within TCPDUMP when it sees STP bridge packets from a CISCO express 500 switch. 15:54:33.774469 802.1d unknown version 15:54:35.787483 802.1d unknown version 15:54:37.801500 802.1d unknown version 15:54:39.372508 00:0d:02:83:7f:18 > 09:00:07:ff:ff:ff snap I (s=0,r=3,P) len=42 15:54:39.766503 00:14:f1:8f:68:01 > 01:00:0c:cc:cc:cc snap I (s=0,r=6,C) len=331 15:54:39.814516 802.1d unknown version 15:54:40.111524 IP 192.168.1.105.2082 > 192.168.1.99.sip-tls: P 22380:22932(552) ack 21013 win 14992 <nop,nop,timestamp 102751459 158489> 15:54:40.128011 IP 192.168.1.99.sip-tls > 192.168.1.105.2082: P 21013:21365(352) ack 22932 win 8406 <nop,nop,timestamp 159228 102751459> 15:54:40.129561 IP 192.168.1.105.2082 > 192.168.1.99.sip-tls: . ack 21365 win 14992 We had previously had a CS410 rebooting, and STP was disabled along with several other QOS improvements and that CS410 suddenly stabilized. A new installation is restarting daily without notice. We have a full 78MB Wireshark Trace for dissection. However, we cannot leave this configuration running for further testing, We will now disable STP on this switch and we fully expect to see the restarts stop... The Cisco Switch also shows about 200 error packets on the port with the CS410, And the switch logs a variety of malformed packets on the CS410 port The Switch in FULL auto has negotiated 100MB Half Duplex.... Quote Link to comment Share on other sites More sharing options...
Vodia PBX Posted June 26, 2009 Report Share Posted June 26, 2009 With a reasonable amount of certainty we believe the CS410 will consistently crash when it sees 802.1D spanning tree packets.The CS410 creates the following within TCPDUMP when it sees STP bridge packets from a CISCO express 500 switch. Whow. There are some rumors that there will be a newer kernel available for the Comcerto platform. The current one is already a few years old... Maybe a new kernel also helps stabilizing this. And it would also help getting IPv6 working on this system. Quote Link to comment Share on other sites More sharing options...
andrewgroup Posted June 26, 2009 Author Report Share Posted June 26, 2009 Whow. There are some rumors that there will be a newer kernel available for the Comcerto platform. The current one is already a few years old... Maybe a new kernel also helps stabilizing this. And it would also help getting IPv6 working on this system. To confirm these findings, configure a layer switch for STP and plug it into you lan and have it generate these stp packets. We saw this behavior on a Linksys SRW208P, and now a CISCO express 500. In the rarest of cases, we may have to put a router on the local lan to segment this traffic if our hunch is correct. Quote Link to comment Share on other sites More sharing options...
andrewgroup Posted June 29, 2009 Author Report Share Posted June 29, 2009 FYI, so far our suspicions are correct and we see no random restarts now that the CS410 is not exposed to STP packets. What we are seeing from many vendors is that STP is enabled by default so as to protect against loops. So we advise to be certain to disable STP on all switch ports if possible or take you chances with a dump switch. (A router with SIP/port control is recommended if you have no access to the LAN switch) Cheers - AndrewGroup Quote Link to comment Share on other sites More sharing options...
pbx support Posted June 30, 2009 Report Share Posted June 30, 2009 FYI,so far our suspicions are correct and we see no random restarts now that the CS410 is not exposed to STP packets. What we are seeing from many vendors is that STP is enabled by default so as to protect against loops. So we advise to be certain to disable STP on all switch ports if possible or take you chances with a dump switch. (A router with SIP/port control is recommended if you have no access to the LAN switch) Cheers - AndrewGroup Thanks for the research & the tip!!! Quote Link to comment Share on other sites More sharing options...
andrewgroup Posted June 30, 2009 Author Report Share Posted June 30, 2009 5 days after the changes we get a restart. (Better than once a day, but still no client can tolerate that.) We see RIP V1 on the local LAN side of the ISP router and that's now turned off and we'll evaluate the results. what would be helpful, would be a small a 5 minute CRON job to gather some stats about the interface, memory, and cpu, that might provide a clue by writing to a series of files to a file to give us an indication of what might be causing a memory leak. Any word on the new Kernel or another update on the PBX or MSP file? (We don't think any of the SNMP values help much.) any thoughts? Quote Link to comment Share on other sites More sharing options...
andrewgroup Posted July 1, 2009 Author Report Share Posted July 1, 2009 We see RIP V1 on the local LAN side of the ISP router and that's now turned off and we'll evaluate the results. The results are the random restarts continue. We'll open a paid support call for guidance and plan to fully replace the hardware with a new unit too. Quote Link to comment Share on other sites More sharing options...
andrewgroup Posted July 17, 2009 Author Report Share Posted July 17, 2009 Sorry for the late Update, but the replacement CS410 has not failed in the last 14 days. Perhaps our original analysis of TSTP packets causing troubles wasn't totally the issue, but removing RSTP did extend the restart issues by 4 days, but the restarts continued but the new CS410 has been stable for 15 days... FYI... This was as new in the BOX cs410 that failed..... So the old adage, if things are going to fail, they will in the first 30 days... Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.