IPsec fails to renegotiate after loss of a peer
-
im using the latest 2.0 snapshot
Don't. That's not going to be stable. Pretty sure the 7.2/2.0 builds still use NAT-T which has renegotiation issues, and the 8 snapshots likely don't have a proper ipsec-tools either.
-
@cmb:
im using the latest 2.0 snapshot
Don't. That's not going to be stable. Pretty sure the 7.2/2.0 builds still use NAT-T which has renegotiation issues, and the 8 snapshots likely don't have a proper ipsec-tools either.
damn… well now that i think of it i really only switched to check out the new gui and multiple dyndns accounts for each of my wans, so ipsec is much more stable and up-to-date in the RC versions?, i would have figured newer code with more features was included in 2.0 for vpns specifically ipsec
-
-
Ok, I enabled DPD (60 sec) and I believe that this fixed part of the problem. A couple of my tunnels stayed up overnight or at least reconnected. One of the tunnels wouldn't reconnect though until I deleted the numerous SADs and restarted raccoon. (Actually a second tunnel stopped after restarting raccoon and I again had to delete the SAD but not restart raccoon) It appears that some of the SADs were not getting dropped properly. The IPSec status showed that the tunnels were up. It may have coincided with the lifetime setting. I think that I may change my lifetimes to a shorter time frame so that I can try to duplicate the behavior. Below are some of the error messages.
racoon: ERROR: fatal INVALID-SPI notify messsage, phase1 should be deleted.
Sep 29 12:23:18 racoon: ERROR: fatal INVALID-SPI notify messsage, phase1 should be deleted.
Sep 29 12:22:48 racoon: ERROR: fatal INVALID-SPI notify messsage, phase1 should be deleted.
Sep 29 12:22:41 racoon: [VPN Name]: INFO: IPsec-SA established: ESP MyWanIPxx.xx[0]->RemoteIPxx.xx[0] spi=2670541922(0x9f2d3c62)
Sep 29 12:22:41 racoon: [VPN Name]: INFO: IPsec-SA established: ESP RemoteIPxx.xx[0]->MyWanIPxx.xx[0] spi=37322644(0x2397f94)
Sep 29 12:22:40 racoon: [VPN Name]: INFO: respond new phase 2 negotiation: MyWanIPxx.xx[0]<=>RemoteIPxx.xx[0]Background: The tunnels are static tunnels to Netopia routers. One tunnel to each site or router.
Aggressive mode is used. For the Keep Alive I am using the remote Lan address of the Netopia router.After I duplicate the problem, I will provide more log info. Most was overwritten before I thought of copying it.
-
I would have figured newer code with more features was included in 2.0 for vpns specifically ipsec
Not in this case. Even so, newer code and more features don't usually translate to more stability, especially in an alpha release :)
-
@bkm:
Ok, I enabled DPD (60 sec) and I believe that this fixed part of the problem. A couple of my tunnels stayed up overnight or at least reconnected. One of the tunnels wouldn't reconnect though until I deleted the numerous SADs and restarted raccoon. (Actually a second tunnel stopped after restarting raccoon and I again had to delete the SAD but not restart raccoon) It appears that some of the SADs were not getting dropped properly. The IPSec status showed that the tunnels were up. It may have coincided with the lifetime setting. I think that I may change my lifetimes to a shorter time frame so that I can try to duplicate the behavior. Below are some of the error messages.
racoon: ERROR: fatal INVALID-SPI notify messsage, phase1 should be deleted.
Sep 29 12:23:18 racoon: ERROR: fatal INVALID-SPI notify messsage, phase1 should be deleted.
Sep 29 12:22:48 racoon: ERROR: fatal INVALID-SPI notify messsage, phase1 should be deleted.
Sep 29 12:22:41 racoon: [VPN Name]: INFO: IPsec-SA established: ESP MyWanIPxx.xx[0]->RemoteIPxx.xx[0] spi=2670541922(0x9f2d3c62)
Sep 29 12:22:41 racoon: [VPN Name]: INFO: IPsec-SA established: ESP RemoteIPxx.xx[0]->MyWanIPxx.xx[0] spi=37322644(0x2397f94)
Sep 29 12:22:40 racoon: [VPN Name]: INFO: respond new phase 2 negotiation: MyWanIPxx.xx[0]<=>RemoteIPxx.xx[0]Background: The tunnels are static tunnels to Netopia routers. One tunnel to each site or router.
Aggressive mode is used. For the Keep Alive I am using the remote Lan address of the Netopia router.After I duplicate the problem, I will provide more log info. Most was overwritten before I thought of copying it.
The full log is in /var/log/ipsec.log, and you can view it by executing the command: clog /var/log/ipsec.log
It may have just been gone from the GUI, which only shows a limited number of lines. Also, it would help to show the logs in normal order, not reverse order. If you have the reverse order box checked on Status > System Logs, Settings tab, uncheck it and save, then copy/paste the logs. There was an old bug that caused the IPsec logs to ignore this setting, but it was fixed before the snapshot you said you were running. You may also want to update to the most recent snapshot to be sure you really have the most current updates.
One more thing: When testing these settings, be sure to stop and restart racoon after making your changes, to be on the safe side and to be sure the SAD and SPD are clear.
-
One more thing: When testing these settings, be sure to stop and restart racoon after making your changes, to be on the safe side and to be sure the SAD and SPD are clear.
sorry if this sounds ignorant but do you mean delete all the SPD and SAD entries?
-
One more thing: When testing these settings, be sure to stop and restart racoon after making your changes, to be on the safe side and to be sure the SAD and SPD are clear.
sorry if this sounds ignorant but do you mean delete all the SPD and SAD entries?
Stopping and restarting racoon should do this. If you want to do it by hand, run:
setkey -F
setkey -F -P -
Thanks for the advice. I knew the logs had to be somewhere. I will also restart raccoon from now on after changes are made.
"sorry if this sounds ignorant but do you mean delete all the SPD and SAD entries?"
In the GUI, I would go to Status-IPSec, then the SAD tab and delete the entries that corresponded to the tunnel that I was having problems with. Normally, I see two entries for each tunnel, MyWanIP to RemoteWanIP and RemoteWanIP to MyWanIP. When I was having problems, this area would get filled with entries where it was trying to connect. There is an X beside each one where they can be deleted. Restarting raccoon also deletes them though. (I was experimenting a little when deleting manually) I did not change anything on the SPD tab.
I did change the lifetime setting to 1800 and 3600, on two of the tunnels so that I could see if they came back up after an expiration without waiting until tomorrow. They did come up this time. If it is down in the morning, I will post a few log entries.
-
One of the tunnels went down again last night. Attached is my IPSec log. It was working at 17:00 on 9/29/09 and was not working at the end of the log on 9/30/09. The log was modified to remove the IP addresses. My Wan IP address is listed as MyWanIP. The remote Wan IP of the tunnel that went down is listed as RemoteWanIPTunnel1. Any advice welcome. Thanks.
-
Well, the tunnel came back up without any action from me other than I did try to ping the site twice. The first time it timed out. An hour later I tried and got an immediate reply. For the Keep Alive address, I am using the remote LAN IP. Should I be using the remote Wan IP instead?
-
I ran a ping test on my tunnels last night to see how long and how often they were down. The test ran for about 15 hours. Tunnel 1 went down 6 times averaging 30-35 minutes each time. Tunnel 2 went down 3 times also averaging 30-35 minutes each time. Tunnel 3 stayed up the whole time. I didn't count the times where the tunnel was only down for about a minute, however this would also be a problem for our VPNs. I cannot verify that times under a couple minutes were not due to a line problem since I alternated my pings between sites. The range for the downtime was between 25 and 45 minutes. Most occurrences appeared to be at 30-35 minutes.
My lifetime settings for tunnel 1 and 2 are at 1800 (30 min) for phase 1 and 3600 (60 min) for phase 2. My lifetimes for tunnel 3 are 28800 and 86400. Tunnel 3's phase 2 lifetime did not expire during the test.
Attached is a partial IPSec log. Tunnel 1 went down sometime between 5:56 and 6:01. It came back up between 6:41 and 6:47 in case someone wants to look. The clocks between the two machines could be slightly off.
I will try to upgrade again in the next day or so and test again, but I am not very hopeful at this point.
Any suggestions are welcome.
Thanks -
I have noticed similar behavior with a pfsense<–>Linksys VPN.
Basically, at some unspecified time, the tunnel appears to die, and stays down for the time the phase 2 lifetime (for me).
I've seen the same behavoir without pfsense in the mix though, using just the Linksys VPN routers (which I am replacing with pfsense boxes, slowly). -
I'm making some progress on my problem, but I am not going to trust it for production yet. I did have a tunnel stay up all night. I am posting this in case it helps someone else. I don't profess to completely understand all of the intricacies of IPSec, but I do have a basic understanding of what happens.
One of the problems with IPSec is different understandings and implementations of the standard. Because the standard is not clear on some issues such as re-keying, different vendors choose different methods.One of my problems was related to "dangling phase 2 SAs". The option on the other end of my tunnels were set to Not allow dangling phase 2 SA's. There is no GUI option for this in pfsense nor is there any option related to either using the newest SAs immediately or to wait until they expire. Either of these can cause a tunnel to go down after the timeout expires.
Pfsense is not wrong in its method, it just has not been made clear in the documentation which method works for it.
I changed one of my tunnels to "Allow Dangling Phase 2 SAs" and it has made a huge difference. This will not help those with a pfsense to pfsense tunnel, but for those who have this option on one end of their tunnel, they should experiment with it.
If either of these options are available in a config file somewhere, I would be interested in knowing.
Thanks
-
Actually, there is an option for using the newest or older SAs. I haven't found anything related to dangling phase 2 SAs though.
-
Has there been any progress with this issue is 1.2.3RC3? I am setting up a new router and heard on the mailing list that not much will change between RC3 and the final release. I am currently using 1.2.2 and everything works great on most of my installations. The major feature I want from 1.2.3 is the ability to make changes to one IPSEC tunnel and not restart each of the 30 tunnels running.
-
Has there been any progress with this issue is 1.2.3RC3? I am setting up a new router and heard on the mailing list that not much will change between RC3 and the final release. I am currently using 1.2.2 and everything works great on most of my installations. The major feature I want from 1.2.3 is the ability to make changes to one IPSEC tunnel and not restart each of the 30 tunnels running.
It appears to be resolved now. I've not had any trouble with my usually-volatile tunnels that used to give me no end of grief making me restart racoon all the time. It's been working like a dream with RC3.
-
Has there been any progress with this issue is 1.2.3RC3?
It's been resolved since NAT-T was removed over a month ago.
-
Thanks to both of you for the information. I already put RC3 on the new router but I was starting to second guess myself because of some other issues that may be unrelated. Hearing that these problems are at the very least not nearly as common gives me the confidence to continue with the project.
-
I'm not sure if I would consider this resolved… I'm running the current 1.2.3-RC3 build and having plenty of IPsec issues that fit this bill. So far, enabling DPD (30 Sec) and 'Prefer old IPsec SA's' seem to have resolved it. It's been ~20 hours and no tunnels have gone down, but I'll give it another day before I feel like it is stable. In any case, this is more of a work around than a resolution.
