High tunnel count (Trac #926)

[str4d]

I had an unexpectedly high tunnel count in the latest version of i2p. The logs mention low bandwidth availability, but since i2p was the only thing running i doubt that this is in fact the case. A soft restart fixed it fine.

Migrated from https://trac.i2p2.de/ticket/926

{
    "status": "assigned", 
    "changetime": "2016-08-05T15:05:21", 
    "description": "I had an unexpectedly high tunnel count in the latest version of i2p. The logs mention low bandwidth availability, but since i2p was the only thing running i doubt that this is in fact the case. A soft restart fixed it fine.", 
    "reporter": "guest", 
    "cc": "HungryHobo", 
    "resolution": "", 
    "_ts": "1470409521135942", 
    "component": "apps/plugins", 
    "summary": "I2P-Bote: High tunnel count", 
    "priority": "minor", 
    "keywords": "I2P-Bote hang", 
    "version": "0.9.5", 
    "parents": "", 
    "time": "2013-04-28T21:39:27", 
    "milestone": "", 
    "owner": "str4d", 
    "type": "defect"
}

[str4d]

Trac update at 20130428T21:40:16:

• guest changed attachment from "" to "logs.txt"
• guest commented:

wrapper log file from the bug

[str4d]

Trac update at 20130428T21:40:40: guest changed attachment from "" to "17.png"

[str4d]

Trac update at 20130429T12:55:20: zzz commented:

You had no peers at all. I2P couldn't connect to anybody. Any idea why?

232 client tunnels but no peers - so they were zero-hop tunnels.

Any clues from the router logs?

[str4d]

Trac update at 20130429T16:56:20:

• guest changed attachment from "" to "log-router-1.txt"
• guest commented:

I've trimmed 5 hours worth of similar messages to reduce filesize

[str4d]

Trac update at 20130429T19:30:53: guest commented:

I see no reason for i2p to be unable to connect. Was the connection error at 5:55 or had it occurred some time previously and only caused a problem at that time? A different install of i2p was unaffected by any issues that did occur.

[str4d]

Trac update at 20130501T13:21:55: zzz commented:

OK, the job lag got up over 1440000 (20 minutes or so) so it had completely exploded. Couldn't even build a zero-hop exploratory tunnel... but that didn't start until 6:00.

We used to have bugs where a connectivity problem caused high CPU... but those should be fixed... and you said there weren't any real internet issues. So I think the high job lag causes peers to go to 0.

I see you're running i2p-bote... I don't know the version numbers, but for most of last year bote would cause a router to eventually explode... it was fixed in the last couple months? Search on here for tickets.

Or maybe that bote-version-checker stuff caused the problem?

What's your i2p-bote version?

[str4d]

Trac update at 20130501T13:45:27: zzz changed cc from "" to "HungryHobo"

[str4d]

Trac update at 20130612T12:08:50:

• zzz changed component from "router/transport" to "apps/plugins"
• zzz changed owner from "zzz" to "HungryHobo"
• zzz changed status from "new" to "assigned"

[str4d]

Trac update at 20140327T23:20:33:

• str4d changed keywords from "" to "i2pbote"
• str4d changed parents from "" to ""

[str4d]

Trac update at 20150103T12:29:11:

• str4d changed keywords from "i2pbote" to "I2P-Bote hang"
• str4d changed milestone from "0.9.6" to ""

[str4d]

Trac update at 20160805T15:05:21:

• zzz changed owner from "HungryHobo" to "str4d"
• zzz changed summary from "High tunnel count" to "I2P-Bote: High tunnel count"