<<< Chronological Index >>>    <<< Thread Index >>>

Re: FYI: created symlinks tt01 -> geuzenveld for old ROOT data


On Mon, 4 Dec 2000, Rene Wilhelm wrote:

> Up untill the last days of 1999 the primary name (as configured with
> "hostname") for tt01 was geuzenveld.ripe.net; consequently, all files
> created by TTM software had the word 'geuzenveld' in it. Around the
> end of 1999, we decided to make tt01 the one and only name for this
> testbox to reduce the number of exeptions in the naming scheme
> (which is: box with id=XY gets hostname ttXY.ripe.net).
> 
> However,  since the configuration "DB" doesn't know of this change and
> returns tt01.ripe.net when asked for the boxes which were online in 1999,
> plots-on-demand would not find any data for this box for the times it went
> by the name of geuzenveld.ripe.net.
> 
> Fotis suggested renaming all existing *geuzenveld.ripe.net* data files
> to *tt01.ripe.net*; however, as raw data and ROOT data files already
> made it to tape with the name geuzenveld and are recorded like that
> in the online tape contents archives, I don't think that's a good idea:
> any reprocessing, refetching files from tape will bring back the
> old name. Instead I've now created symlinks in the appropriate 1999
> root_data directory trees, such that plots-on-demand will find all
> data for tt01.ripe.net.
> 
> It would probably be nice if sometime in future we could avoid the
> convolution of testbox ids and names (e.g. it's pretty confusing if
> tt98.ripe.net plots appear under www/test-traffic/Plots/tt05),
> but for now it seems best to follow the plain and simple rule
> hostname=tt<BoxID>.ripe.net for all machines participating in 
> the measurement chain. Also: once datataking started, a hostname
> should ALWAYS be the same for the duration of a box' active TTM life.

Agreed.  

A long time ago, we also agreed that TB names should not be recycled, in
order to avoid confusion when analyzing old data.  tt05/tt06 now seem to
violate that rule (they were machine in Groningen and Dortmund early '98).
I have no particular feeling about this for test-machines (you'll probably
won't include these in any long term things), but if/when the
test-machines become permanent additions to the chain, they should have a
unique number.  (In fact, tt03 was a test-box sometime, for 3 days at the
Park Plaza during RIPE29)



Henk

------------------------------------------------------------------------------
Henk Uijterwaal                    Email: henk.uijterwaal@ripe.net
RIPE Network Coordination Centre     WWW: http://www.ripe.net/home/henk
Singel 258                         Phone: +31.20.535-4414,  Fax -4445
1016 AB Amsterdam                   Home: +31.20.4195305
The Netherlands                   Mobile: +31.6.55861746  
------------------------------------------------------------------------------

A man can take a train and never reach his destination.
                                               (Kerouac, well before RFC2780).






<<< Chronological Index >>>    <<< Thread Index >>>