[Dxspider-support] GIT?

Dirk Koopman djk at tobit.co.uk
Sun May 31 12:14:46 CEST 2020


On 31/05/2020 10:26, Danilo Brelih via Dxspider-support wrote:
> Dirk Koopman via Dxspider-support je 31.5.2020 ob 0:28 napisal:
>> I suggest you do a 'git checkout e42d9225'
>>> and restart the node.
>
> Thanks all for suggestions. After restarting the node there is no more 
> extreme high cpu load but I  still have a mess here. Spider build is 
> now on 244!
>
> DX Spider Cluster version 1.57 (build 244 git: (HEAD/e42d9225[r]) on 
> Linux
> Copyright (c) 1998-2020 Dirk Koopman G1TLH
>
> When git pull again I get
>
> You are not currently on a branch.
> Please specify which branch you want to merge with.
>

Just: git checkout mojo

Which will get you I still don't understand why you should have the CPU 
issues when others don't. They have problems getting the update working 
and don't seem to be able understand when to do apt installs or even 
what that means. But it seems to work fine once they have the missing 
packages loaded.

The big ticket difference is that (notwithstanding the remarks in the 
Changes file about ditching Storable) between where you are and the HEAD 
is that the Spot::search has been rewritten to leverage the Filtering 
system (and language) to do sh/dx. There should be no observable 
difference between old sh/dx and new sh/dx except that you can now use 
brackets, and, or and not.  But I am aware that there may be some 
optimisations that I make to speed it up (there is an experimental one 
in operation which might be a issue, in that I am using the unix 'tac' 
command  to read files backwards which is much faster for the general 
case of sh/dx with no parameters other than a count, but (only slightly) 
slower doing sh/dx <callsign> on a rare call going back 100 days). It 
can be disabled, by parameter, if that's what you think is a problem.

How many users/nodes do you have?

Still only on very intermittent internet...

Dirk




More information about the Dxspider-support mailing list