Sorry for my late reply I was out of town most of last week. The newest build seems to have fix the problem with it almost always being overloaded, but it still is getting overloaded by this by a lot. It will get overloaded stay that way for about 5 mins then fall below the threshold and start accepting connections again for about 5 mins till it shoots back up and is overloaded again and restarts the cycle.
Its been up for 10.5 hours, I had to copy the info out piece by piece to get it so sorry if it looks different (the firewall here doesn't like port 8888 so I have to use a telnet prompt to get to my machine and dump the info with a program I wrote...) Current routingTime: 0ms Current messageSendTimeRequest: 4741ms Pooled threads running jobs: 71 (47.3%) Pooled threads which are idle: 14 Reason for refusing connections: avgMessageSendTimeRequest(4741.311) > successfulSendTimeCutoff(3000.000) Current estimated load for QueryReject purposes: 47% Current estimated load for rate limiting: 237% Reason for load: Load due to thread limit: 47.3% Load due to routingTime: 3.3% = 100ms / 3000ms <= overloadLow (100%) Load due to messageSendTimeRequest: 237% = 4741ms / 2000ms > overloadLow (100%) Load due to expected inbound transfers: 3.4% because 143028.4384083343 req/hr * 0.0010589149440000006 (pTransfer) * 291384.0 bytes = 44131549 bytes/hr maxInputBytes/minute = 19509360 (max observed bytes per minute) * 60 * 1.1 = 1287617760 bytes/hr target Estimated external pSearchFailed (based only on QueryRejections due to load): 0.499004276735748 Current estimated requests per hour (based on last 10 mins): 42216.705203521495 Current global quota (requests per hour): 40460.32455491089 Highest seen bytes downloaded in one minute: 19509360 Current outgoing request rate: 143028.4384083343 Current probability of a request succeeding: 0.4% Current target (best case single node) probability of a request succeeding: 6.5% -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of [EMAIL PROTECTED] Sent: Friday, May 28, 2004 5:59 PM To: Findley, Matthew; [EMAIL PROTECTED] Subject: Re: [freenet-support] messageSendTimeRequest Importance: Low On Fri, May 28, 2004 at 01:13:01PM -0400, [EMAIL PROTECTED] wrote: > My node is almost constantly overloaded because messageSendTimeRequest exceeds > successfulSendTimeCutoff by a significant amount. The messageSendTimeRequest likes > to hang around 4000ms. Now I've observed that the messageSendTimeRequest seems to > be tied the amount of data waiting to be transferred so I've tried to tweak the > settings to reduce the amount of data that is going out of my node. Reducing the > number of max connections, reducing the chance of cache, and other little things. > But they don't seem to be able to bring it down. The only thing I can figure is > that my node just looks too good to pass up (I have a really good upstream), and the > other nodes just really like to request from my node and end up overloading me. > So is there anything I can do to reduce the load? Okay, this is wierd. You running stable, I assume? Can you send me the output of your General infolet? http://127.0.0.1:8888/servlet/nodeinfo/performance/general How long has the node been running? -- Matthew J Toseland - [EMAIL PROTECTED] Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so. _______________________________________________ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
