trafficserver-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Igor Galić <i.ga...@brainsware.org>
Subject Re: Upgrade procedure.
Date Thu, 21 Nov 2013 12:06:31 GMT
----- Original Message -----

> On 21.11.2013 11:54, Igor Galić wrote:

> > ----- Original Message -----
> 

> > > Anyway, I don't understand why I can't use the same cache when I switch
> > > over
> > > to newly compiled version?
> > 
> 
> > > I've checked the object with cache inspector and found, that the key (not
> > > url-key) is different for same object under 4.0.1 and 4.1.1, but "first
> > > key"
> > > is the same:
> > 
> 

> > > ATS 4.0.1
> > 
> 
> > > Volume #1 - store='/dev/sda'
> > 
> 
> > > first key 409542BD429764BEE60B0610B8924C4D
> > 
> 
> > > key 6BA7E5696E9A9E7A1E05212E5264D3C4
> > 
> 
> > > sync_serial 10836
> > 
> 
> > > write_serial 388912
> > 
> 
> > > header length 2480
> > 
> 
> > > fragment type 1
> > 
> 
> > > No of Alternates 1
> > 
> 

> > > ATS 4.1.1
> > 
> 
> > > Volume #1 - store='/dev/sda'
> > 
> 
> > > first key 409542BD429764BEE60B0610B8924C4D
> > 
> 
> > > key 34CEA58AC5FBA6D240C484307DE4C315
> > 
> 
> > > sync_serial 10837
> > 
> 
> > > write_serial 388912
> > 
> 
> > > header length 2480
> > 
> 
> > > fragment type 1
> > 
> 
> > > No of Alternates 1
> > 
> 

> > > Should I copy anything else but *.config files from previous 4.0.1 to
> > > 4.1.1
> > > to make 4.1.1 found cached objects?
> > 
> 

> > I would copy the *changes* you made to your .config files.
> 

> > The issues with the key changing is a little off putting.
> 

> > Do you have any means of verifying verify that ATS serves
> 
> > the correct objects from cache when run under 4.1.1?
> 

> I'm not sure what you mean... When run 4.1.1 all previously cached objects
> under 4.0.1 are MISS, these objects downloading from parent, and then they
> HIT again
Well, this is bad. This is especially bad because we advertised this release as 4.x compatible.

I'm very surprised none of our power-users (Yahoo, et al) noticed this. 

Currently I'm giving the code a stare-down to see what changed that would prompt the 
on-disk cache to become incompatible. 

-- i 
Igor Galić 

Tel: +43 (0) 664 886 22 883 
Mail: i.galic@brainsware.org 
URL: http://brainsware.org/ 
GPG: 8716 7A9F 989B ABD5 100F 4008 F266 55D6 2998 1641 

Mime
View raw message