Re: Configured hibernation does not work after restart in an interval?

[Joerg Maschtaler <mali@xxxxxxxxxxxx>]

Hi Roger,

First, i used the wrong logfile. Sorry.
The notice from the logfile was from another server.

Nevertheless the limit was exceeded on the server i meant.

Roger Dingledine schrieb am 05.08.2006 16:20:
> On Thu, Aug 03, 2006 at 01:52:17PM +0200, Joerg Maschtaler wrote:
>> i use Tor 0.1.1.22 as a server.
>> The limit for traffic which i defined with option AccountingMax was
>> exceeded for both directions in the given interval.

[Wrong notice deleted]

> Is your DataDirectory writable by your Tor process? 

Yes.

> Is your Tor process creating and upating a file called "bw_accounting"
> in the DataDirectory?

Yes.

> What OS, and did you install from package or source, and did you 
> change anything? Any other warnings in your Tor log?

OS: SuSE 9.3
Tor source was compiled. In addition to the installations instructions i
made a "make install".
File torrc was untouched.


From June to now the logfile shows several events:

06/06/12 - Restart of 0.1.1.21
---snip---
Jun 12 10:36:37.125 [notice] accounting_set_wakeup_time(): Configured
hibernation.  This interval began at 2006-06-01 00:00:00; the scheduled
wake-up time was 2006-06-01 00:00:00; we expect to exhaust our quota for
this interval around 2006-07-01 00:00:00; the next interval begins at
2006-07-01 00:00:00 (all times local)
---snap---


06/07/01 - Message of accounting
---snip---
Jul 01 00:00:00.071 [notice] accounting_set_wakeup_time(): Configured
hibernation.  This interval began at 2006-07-01 00:00:00; the scheduled
wake-up time was 2006-07-01 00:00:00; we expect to exhaust our quota for
this interval around 2006-08-01 00:00:00; the next interval begins at
2006-08-01 00:00:00 (all times local)
---snap---


06/06/10 - Installation and start of Tor 0.1.1.22
---snip---
Jul 10 20:09:28.798 [notice] accounting_set_wakeup_time(): Configured
hibernation.  This interval began at 2006-07-01 00:00:00; the scheduled
wake-up time was 2006-07-09 04:06:52; we expected to exhaust our quota
for this interval around 2006-07-09 12:41:52; the next interval begins
at 2006-08-01 00:00:00 (all times local)
---snap---


06/06/11 - Clean shutdown and restart due to an amendment of AccountingMax.
---snip---
Jul 11 00:10:11.713 [notice] hibernate_begin(): Interrupt: will shut
down in 10 seconds. Interrupt again to exit now.
Jul 11 00:10:22.009 [notice] consider_hibernation(): Clean shutdown
finished. Exiting.
Jul 11 00:11:30.159 [notice] accounting_set_wakeup_time(): Configured
hibernation.  This interval began at 2006-07-01 00:00:00; the scheduled
wake-up time was 2006-07-01 00:00:00; we expect to exhaust our quota for
this interval around 2006-08-01 00:00:00; the next interval begins at
2006-08-01 00:00:00 (all times local)
---snap---


06/08/01 - Message of accounting
---snip---
Aug 01 00:00:00.471 [notice] accounting_set_wakeup_time(): Configured
hibernation.  This interval began at 2006-08-01 00:00:00; the scheduled
wake-up time was 2006-08-01 00:00:00; we expect to exhaust our quota for
this interval around 2006-09-01 00:00:00; the next interval begins at
2006-09-01 00:00:00 (all times local)
---snap---



Aside from the messages above the are two kinds of warnings:

5 times till 21th June:
---snip---
Jun 20 22:59:03.556 [warn] crypto error while checking RSA signature:
data greater than mod len (in rsa routines:RSA_EAY_PUBLIC_DECRYPT)
Jun 20 22:59:03.578 [warn] crypto_pk_public_checksig_digest(): Invalid
signature
---snap---

and several times:
---snip---
Jul 01 17:29:18.158 [warn] connection_connect(): Error creating network
socket: No buffer space available
---snap---
till 1st July 18:08.



The limit was approximately exceeded on 29th July.
Maybe the change of AccountingMax or the upgrade of Tor on 06/06/10
could be a reason?


Thx in advance,
Joerg