Expand my Community achievements bar.

Dive into Adobe Summit 2024! Explore curated list of AEM sessions & labs, register, connect with experts, ask questions, engage, and share insights. Don't miss the excitement.
SOLVED

AEM6 SP1 - Frequent Re-Indexing

Avatar

Level 2

We have the environment Linux with AEM6 SP1. There is an external integration that brings some content (product pages and associated images) from a 3rd party service. This service is called from author via a scheduler to create relevant JCR nodes. The odd behaviour that we observe that which each such load (it runs weekly), the repository does a full reindexing of content and that makes author instance unavailable for some a good time period.

Question:

[1] Is there a way to identify what actually triggers this full reindexing and possible solution to resolve this? The data that comes through the service is not very heavy comparison to what we already have in the repo and thus doing entire reindexing seems unwarranted.

I am also trying to follow the hep article https://helpx.adobe.com/experience-manager/kb/performance-tuning-tips.html -- > especially "Fins Slow Queries" using qla.jar but not so encouraging result yet. Any suggestion, on similar kind of issues would be very helpful.

Thanks...

1 Accepted Solution

Avatar

Correct answer by
Employee Advisor

Hi,

What makes you think, that the repository is doing a "full reindex"? And what do you mean with "full reindex"? With Oak there is no single index anymore, but AEM 6.0 brings ootb a number of specific indexes. Which of these indexes is affected?

kind regards,
Jörg

View solution in original post

5 Replies

Avatar

Correct answer by
Employee Advisor

Hi,

What makes you think, that the repository is doing a "full reindex"? And what do you mean with "full reindex"? With Oak there is no single index anymore, but AEM 6.0 brings ootb a number of specific indexes. Which of these indexes is affected?

kind regards,
Jörg

Avatar

Level 2

Of total repo size of 30 gig, a 3%-4% change through nightly job ingestion causes the authoring to do indexing of content for hours (almost better part of the day. This points to some inconsistencies in either indexes or mechanism of indexing. This is not the behavior if we create a bulk of editorial content. We are trying to check if we have all possible indexes defined to cover different combinations of content being ingested.

Is there any best practices available on how should we go about debugging / analyzing indexes or optimize them for best performances? I guess a lot to be done by us before we engage Adobe and thus seeking any such experiences that y'll might have in this type of issues.

Thanks for the responses.

Avatar

Employee

Hi Anand,

can you post some of the log messages you see for your reindexing issue?

Regards,

Opkar

Avatar

Employee

Hi,

can this issue be replicated in your dev and local instances? Have you debugged your code to see at which point the reindex is triggered? As you suggest this is not normal behaviour, have you opened a daycare ticket?

Regards,

Opkar

Avatar

Level 2

Unfortunately, it has not yet been replicated in DEV. We are doing following:

1) Looking at external data that get ingested and check the oak-indexes to ensure whether all content types /  properties are covered

2) We are also monitoring all queries and its responses

I guess Adobe would need these info for their analysis as well.

Thanks

Anand