venti survey

venti survey

[Russ Cox]

I am curious about the relative sizes of people's venti
servers, how much data they have stored, and how
quickly they are filling.

Please run

	hget http://my-venti-url/storage >/tmp/v
	hget http://my-venti-url/index >>/tmp/v

	mail -s 'venti survey' rsc@swtch.com </tmp/v

In a week or so I will post a summary.

Thanks.
Russ

Re: [9fans] venti survey

[Adrian Tritschler]

[-- Attachment #1: Type: text/plain, Size: 1243 bytes --]

Russ Cox wrote:
> I am curious about the relative sizes of people's venti
> servers, how much data they have stored, and how
> quickly they are filling.
>
> Please run
>
> 	hget http://my-venti-url/storage >/tmp/v
> 	hget http://my-venti-url/index >>/tmp/v
>
> 	mail -s 'venti survey' rsc@swtch.com </tmp/v
>
> In a week or so I will post a summary.

Russ,

If it is of interest to you I've attached the logs of the ventis that
I've run for the last year or two.  In both cases they've been running
on plan9ports, with arenas and indexes held in files in linux filesystems.

The logs are from venti running on two linux systems:

fafnir: venti used for "tinkering" and vbackup of some very small linux
file systems

nidhogg/w-cl28-g34-c: venti for a VMware guest plan9 cpu server, and
vbackup of some linux systems.  (The hostname changed

I put the logging into my crontab ages ago with the intention of
generating my own graphs but never got around to it.

The logs were generated with a very simplistic hourly cron job containing:

	VENTILOG=$HOME/log/venti-$HOSTNAME.log
	$PLAN9/bin/date -n >> $VENTILOG
        $PLAN9/bin/hget http://localhost:1088/storage >> $VENTILOG

> Thanks.
> Russ
	Adrian


[-- Attachment #2: venti.zip --]
[-- Type: application/zip, Size: 199184 bytes --]

Re: [9fans] venti survey

[Adrian Tritschler]

Adrian Tritschler wrote:

...without checking the reply address.

appologies for the noise.

btw, what has happened to the plan9 website?
http://plan9.bell-labs.com/ redirects to http://go.cs.bell-labs.com/ and
only the index page of the wiki exists, the others redirect to 404 errors.

	Adrian

Re: [9fans] venti survey

[Uriel]

>From the wiki page there:  Last modified Tue Apr 9 02:55:52 EDT 2002

In any case, seems that the current site (with the current wiki) is
still up and running at one of the multiple addresses where one has
been able to find it over the years: http://cm.bell-labs.com/plan9/

uriel


On Tue, Jun 24, 2008 at 4:39 AM, Adrian Tritschler <ajft@ajft.org> wrote:
> Adrian Tritschler wrote:
>
> ...without checking the reply address.
>
> appologies for the noise.
>
> btw, what has happened to the plan9 website? http://plan9.bell-labs.com/
> redirects to http://go.cs.bell-labs.com/ and only the index page of the wiki
> exists, the others redirect to 404 errors.
>
>        Adrian
>
>
>

Re: [9fans] venti survey

[Eric Van Hensbergen]

[-- Attachment #1: Type: text/plain, Size: 545 bytes --]

I finally coerced my plan 9 venti to output the stuff.  Should be attached.



On Thu, Jun 19, 2008 at 10:23 AM, Russ Cox <rsc@swtch.com> wrote:
> I am curious about the relative sizes of people's venti
> servers, how much data they have stored, and how
> quickly they are filling.
>
> Please run
>
>        hget http://my-venti-url/storage >/tmp/v
>        hget http://my-venti-url/index >>/tmp/v
>
>        mail -s 'venti survey' rsc@swtch.com </tmp/v
>
> In a week or so I will post a summary.
>
> Thanks.
> Russ
>
>
>

[-- Attachment #2: ibm-plan9-venti.txt --]
[-- Type: text/plain, Size: 2081 bytes --]

index=work
total arenas=395 active=4
total space=212,057,538,560 used=1,970,914,507
clumps=742,362 compressed clumps=609,642 data=4,413,433,588 compressed data=1,924,145,701
index=work version=1 blocksize=8192 tabsize=65536
	buckets=1294445 div=3318
	sect=isect0 for buckets [0,1294445) buckmax=215
arena=areanas00 at index [1048576,537903104)
	arena='areanas00' on /dev/sdC1/arenas at [344064,537214976)
	version=4 created=1172077334 modified=1176872502 mem=sealed disk=sealed
	score=20c2135290b4bacf194b3124421778fc509892ea
	written: clumps=179171 compressed clumps=147810 data=1,094,791,548 compressed data=525,556,018 storage=536,843,791
	indexed: clumps=179171 compressed clumps=147810 data=1,094,791,548 compressed data=525,556,018 storage=536,843,791
arena=areanas01 at index [537903104,1074757632)
	arena='areanas01' on /dev/sdC1/arenas at [537214976,1074085888)
	version=4 created=1176872502 modified=1200504304 mem=sealed disk=sealed
	score=051e949bf081683923cb7bd0ee9cd685452613e4
	written: clumps=228928 compressed clumps=198920 data=1,260,106,548 compressed data=522,409,702 storage=536,832,166
	indexed: clumps=228928 compressed clumps=198920 data=1,260,106,548 compressed data=522,409,702 storage=536,832,166
arena=areanas02 at index [1074757632,1611612160)
	arena='areanas02' on /dev/sdC1/arenas at [1074085888,1610956800)
	version=4 created=1200504304 modified=1205992466 mem=sealed disk=sealed
	score=d4f6b9f431cfc6fbbb2b6f9a20026f0ce850d86b
	written: clumps=236039 compressed clumps=191905 data=1,393,317,760 compressed data=521,968,515 storage=536,838,972
	indexed: clumps=236039 compressed clumps=191905 data=1,393,317,760 compressed data=521,968,515 storage=536,838,972
arena=areanas03 at index [1611612160,2148466688)
	arena='areanas03' on /dev/sdC1/arenas at [1610956800,2147827712)
	version=4 created=1205992466 modified=1214452836
	written: clumps=98224 compressed clumps=71007 data=665,217,732 compressed data=354,211,466 storage=360,399,578
	indexed: clumps=98224 compressed clumps=71007 data=665,217,732 compressed data=354,211,466 storage=360,399,578

Re: [9fans] venti survey

[erik quanstrom]

> total space=212,057,538,560 used=1,970,914,507
> clumps=742,362 compressed clumps=609,642 data=4,413,433,588 compressed data=1,924,145,701

the total used space is less than the amount
of compressed data.  am i reading this incorrectly?

- erik

Re: [9fans] venti survey

[Juan Céspedes]

On Thu, Jun 26, 2008 at 10:18 PM, erik quanstrom <quanstro@quanstro.net> wrote:
> total space=212,057,538,560
> used=1,970,914,507
> clumps=742,362
> compressed clumps=609,642
> data=4,413,433,588
> compressed data=1,924,145,701
>
> the total used space is less than the amount
> of compressed data.  am i reading this incorrectly?

I think so... :)

Juan Cespedes