[Vtigercrm-developers] Email marketing scenario

Błażej Pabiszczak b.pabiszczak at yetiforce.com
Wed Jul 20 07:06:01 GMT 2016


Please answer me the following questions so we can discuss further: 

1. About how much data are we talking about [in round numbers] and best
if it would be with the division into modules.
2. What info do you have to search in the search engine [I suppose not
all info]? Is it enough to define fields, e.g. A, D, E, J for every
module and search by them?
3. Does the search engine need any special search attributes? 

---
Z poważaniem / Regards 

BŁAŻEJ PABISZCZAK 
_Chief Executive Officer_ 
M: +48.884999123
E: b.pabiszczak at yetiforce.com 
-------------------------

YetiForce 3.0 LTS has arrived! Test [1] the latest, most innovative open
source system in the world, and join [2] our community. 

W dniu 2016-07-15 16:19, Alan Bell napisał(a):

> On 15/07/16 14:31, Błażej Pabiszczak wrote: 
> 
>> Whose fault is it that you used a sledgehammer to crack a nut?
> well I was using it to crack some concrete, it isn't designed for the nut situation and I have been quite clear about that all along. 
> 
>> For the search in the system to work efficiently, and quickly, and additionally for it to be configurable, you don't need that much work [2 ~3 days] and for 99% of clients that's enough. Even if that 1% needs more effective or elastic search, then you can expand the search engine by some extra mechanisms, so that it becomes part of the system. The moment when it becomes necessary to use an external mechanism usually happens in case of really large systems, that are used by thousands of users simultaneously, and everything that can relieve the system really matters.
>> But you are entitled to your own opinions, in mine - you don't have a client who required an external search engine.
> trust me, we do! it is not so much the number of users, but the number of records, doing a column like "%query%" query for every column in every table does not scale to large databases (which is what the pre-6.x search did and took >1 minute of maxed CPU to return and killed the system for other users when a search was happening). The % at the start of the like means that it can't use any indexes even if you did index every single column (which is also a bad idea for insert performance and disk space). 
> 
>> You made it this way, so that you can sell it and justify why a basic mechanism like that has to be purchased.
> nah, made it that way because we needed it, and were already using Sphinx in 5.x. Popped it in the marketplace because we had already made it and it might be useful to others. 
> 
>> I have absolutely no problem with that, it's alright for me, just don't say that it couldn't have been done in a different, simpler way, beneficial to all Vtiger users.
> I wouldn't say that at all, I am sure it can be done, but I know that the brute force MySQL query method doesn't scale, and the label only method doesn't find stuff. I did have a look at the full text search facility built into mysql https://dev.mysql.com/doc/refman/5.7/en/fulltext-search.html but that appears to have lots of limitations and doesn't really fit what I wanted it to do.
> How would you go about making search work in vtiger/yetiforce?
> 
> Alan.
> _______________________________________________
> http://www.vtiger.com/
 

Links:
------
[1] https://gitdeveloper.yetiforce.com/
[2] https://github.com/YetiForceCompany/YetiForceCRM
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.vtigercrm.com/pipermail/vtigercrm-developers/attachments/20160720/02451325/attachment-0001.html>


More information about the vtigercrm-developers mailing list