Circuit creation "storms" overwhelming Raspberry Pi?
Hi there, I just joined the mailing list and apologized if this has been discussed before. I did find discussion of a similar issue in January 2013's archive: https://lists.torproject.org/pipermail/tor-relays/2013-January/001809.html It's important to note that I believe I've seen (but didn't save logs) a couple "circuit creation burst" events on my established relay (about 5Mbps, stable, guard, non-exit) which was mostly able to handle it without crashing as it has plenty of RAM and the above-mentioned messages - "Your computer is too slow to handle this many circuit creation requests! Please consider using the MaxAdvertisedBandwidth config option or choosing a m ore restricted exit policy." - appear only with the relay is under load for other reasons AND a large number of circuits are being suddenly created. I wondered if this was some kind of DOS attempt but didn't think much of it because my fast relay continued working fine. However, I've just set up a Raspberry Pi, the 512MB model, as a relay on a slower connection. Here are the relevant settings on this relay: RelayBandwidthRate 130 KB RelayBandwidthBurst 340 KB The Pi has a fairly slow CPU, so I'd occasionally get messages about log deduplication being too slow or something, but didn't think much of it. I finally got the relay up and left it up for over 24 hours. When I woke up this morning it had crashed. Here are the relevant log messages - note the huge jump in number of circuits between 22:35 and 04:35 (maybe I got the Stable flag), then the storm of circuit open requests starting at 05:49. Eventually I believe the Pi ran out of memory and killed the tor process. What's very interesting here is that my fast VPS relay with a RelayBandwidthRate over 5x faster is almost never handling much more than 1000 circuits, so why all of a sudden the demand on the Pi that's advertising a lower bandwidth rate? Mar 17 22:35:00.000 [notice] Heartbeat: Tor's uptime is 1 day 0:00 hours, with 26 circuits open. I've sent 974.13 MB and received 969.92 MB. Mar 18 04:35:00.000 [notice] Heartbeat: Tor's uptime is 1 day 6:00 hours, with 972 circuits open. I've sent 1.61 GB and received 1.59 GB. Mar 18 05:49:44.000 [warn] Your computer is too slow to handle this many circuit creation requests! Please consider using the MaxAdvertisedBandwidth config option or choosing a more restricted exit policy. Mar 18 05:49:44.000 [warn] Failed to hand off onionskin. Closing. Mar 18 05:50:44.000 [warn] Your computer is too slow to handle this many circuit creation requests! Please consider using the MaxAdvertisedBandwidth config option or choosing a more restricted exit policy. [5817 similar message(s) suppressed in last 60 seconds] Mar 18 05:52:30.000 [warn] Your system clock just jumped 101 seconds forward; assuming established circuits no longer work. Mar 18 05:53:51.000 [warn] Your computer is too slow to handle this many circuit creation requests! Please consider using the MaxAdvertisedBandwidth config option or choosing a more restricted exit policy. [1055 similar message(s) suppressed in last 60 seconds] Mar 18 05:55:14.000 [warn] Your computer is too slow to handle this many circuit creation requests! Please consider using the MaxAdvertisedBandwidth config option or choosing a more restricted exit policy. [329 similar message(s) suppressed in last 60 seconds] I'd like to figure out just how much the Raspberry Pi is capable of, because it could be a cheap way to build out the relay network by people who want to donate bandwidth - but of course it needs to be stable, and something about my setup is not. Also: Mar 16 20:55:33.000 [notice] No AES engine found; using AES_* functions. I have no idea if the Broadcom BCM2835 SoC (ARM1176JZF-S CPU) in the Pi has any AES capability, but it'd be great to find out.
I'm also seeing occasional messages like this on the Pi (it never lasts long): 18:13:24 [ARM_NOTICE] Relay resumed 18:13:18 [ARM_NOTICE] Relay unresponsive (last heartbeat: Mon Mar 18 18:13:04 2013) 17:28:43 [ARM_NOTICE] Relay resumed 17:28:38 [ARM_NOTICE] Relay unresponsive (last heartbeat: Mon Mar 18 17:28:25 2013) 14:12:26 [ARM_NOTICE] Relay resumed 14:12:20 [ARM_WARN] Deduplication took too long. Its current implementation has difficulty handling large logs so disabling it to keep the interface responsive. 14:12:20 [ARM_NOTICE] Relay unresponsive (last heartbeat: Mon Mar 18 14:12:06 20 On Mon, Mar 18, 2013, at 01:00 PM, torsion@ftml.net wrote:
I did a lot of tuning on the Raspberry Pi and it's now much, much more stable as a Tor relay, but just now I had another "circuit creation storm." Interestingly, the Pi remained up, and my *router* crashed. I've also seen huge bursts of circuit creation on a relay I run on a VPS, but as it's a much more powerful box it rarely complains (and thus I rarely notice). This many circuits and outbound connections is highly unusual for the small relay I'm running on the Pi. And this behavior definitely occurs in bursts. Is this an outbound DDOS or an attack on Tor itself? If the former (or maybe the latter), is there some way I could perhaps use iptables to temporarily "clamp" the ability to open TCP connections when Tor (or anything on the Pi) opens a number over some threshold in some short period of time? Here's log output (via 'arm') from the relay after my router crashed twice, I went to the admin panel and noted hundreds of outbound connections from my Tor box. Time is America/Los_Angeles. │ 13:55:00 [ARM_NOTICE] Relay unresponsive (last heartbeat: Sat May 4 13:54:14 2013) │ 13:52:25 [WARN] Your computer is too slow to handle this many circuit creation │ requests! Please consider using the MaxAdvertisedBandwidth config option or choosing │ a more restricted exit policy. [404 similar message(s) suppressed in last 60 seconds] │ 13:51:07 [WARN] Your computer is too slow to handle this many circuit creation │ requests! Please consider using the MaxAdvertisedBandwidth config option or choosing │ a more restricted exit policy. [75 similar message(s) suppressed in last 60 seconds] │ 13:50:52 [WARN] Your computer is too slow to handle this many circuit creation │ requests! Please consider using the MaxAdvertisedBandwidth config option or choosing │ a more restricted exit policy. [601 similar message(s) suppressed in last 60 seconds] │ 13:48:39 [WARN] Your computer is too slow to handle this many circuit creation │ requests! Please consider using the MaxAdvertisedBandwidth config option or choosing │ a more restricted exit policy. [99 similar message(s) suppressed in last 60 seconds] │ 13:47:34 [WARN] Your computer is too slow to handle this many circuit creation │ requests! Please consider using the MaxAdvertisedBandwidth config option or choosing │ a more restricted exit policy. [22 similar message(s) suppressed in last 60 seconds] │ 13:46:17 [WARN] Your computer is too slow to handle this many circuit creation │ requests! Please consider using the MaxAdvertisedBandwidth config option or choosing │ a more restricted exit policy. [253 similar message(s) suppressed in last 60 seconds] │ 13:43:47 [WARN] Your computer is too slow to handle this many circuit creation │ requests! Please consider using the MaxAdvertisedBandwidth config option or choosing │ a more restricted exit policy. [1396 similar message(s) suppressed in last 60 │ seconds] │ 13:42:48 [WARN] Your computer is too slow to handle this many circuit creation │ requests! Please consider using the MaxAdvertisedBandwidth config option or choosing │ a more restricted exit policy. [16 similar message(s) suppressed in last 60 seconds] Here's how it crashed my router (blowing ip_conntrack limits is sufficient only to mess up many of my TCP connections, but eventually the router runs out of memory and starts killing processes): May 4 13:51:24 dedmaus user.warn kernel: ip_conntrack: table full, dropping packet. May 4 13:51:24 dedmaus user.warn kernel: ip_conntrack: table full, dropping packet. May 4 13:51:24 dedmaus user.warn kernel: ip_conntrack: table full, dropping packet. May 4 13:51:25 dedmaus user.warn kernel: ip_conntrack: table full, dropping packet. May 4 13:51:29 dedmaus user.warn kernel: NET: 152 messages suppressed. May 4 13:51:29 dedmaus user.warn kernel: ip_conntrack: table full, dropping packet. May 4 13:51:34 dedmaus user.warn kernel: NET: 193 messages suppressed. May 4 13:51:34 dedmaus user.warn kernel: ip_conntrack: table full, dropping packet. May 4 13:51:39 dedmaus user.warn kernel: NET: 227 messages suppressed. ...ad infinitum with the number of messages suppressed per 5 sec increasing until the router crashes. On Mon, Mar 18, 2013, at 06:18 PM, torsion@ftml.net wrote:
Hi torsion, I'm also running a tor relay on a raspberry pi and keep getting these storm creation events which crash the box. You said you made some adjustments to the configs to get a more stable system? Can you please email me copies of the configs or maybe list the changes you made? I'm trying to get my pi stable and then perhaps role out a few more relay nodes to friends and colleagues. Thanks, T On 4 May 2013 22:07, <torsion@ftml.net> wrote:
Hi, Yes. This is absolutely on my to-do list. I've had a family medical emergency and about 2 or 3 other things recently about that level of stress, but BELIEVE me, a strategy for getting a Raspberry Pi to be a rock solid relay is of paramount importance to me. I'm hoping to figure out all the tweaks I did, and then make an alpha version of some iptables rules to defend against the "storms," within the next couple of weeks. Please don't hesitate to bug me about this at my direct email. (The old one is no longer valid, I've dropped pseudonyms on the tor lists.) Thanks much, -Gordon M. Thomas Hand:
participants (3)
-
Gordon Morehouse -
Thomas Hand -
torsion＠ftml.net