Bug 1635685

Summary:	[RFE] Create warning for the users, not to create large sync plans, and rather use smaller batches.
Product:	Red Hat Satellite	Reporter:	Stefan Nemeth <snemeth>
Component:	Documentation	Assignee:	Sergei Petrosian <spetrosi>
Status:	CLOSED WONTFIX	QA Contact:	Melanie Corr <mcorr>
Severity:	medium	Docs Contact:
Priority:	medium
Version:	6.3.3	CC:	adahms, bkearney, daniele, daviddavis, ktordeur, pmoravec, spetrosi
Target Milestone:	Released	Keywords:	FutureFeature, Reopened
Target Release:	Unused
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2018-11-09 15:58:36 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Stefan Nemeth 2018-10-03 13:30:58 UTC

Document URL: 

https://access.redhat.com/documentation/en-us/red_hat_satellite/6.3/html/content_management_guide/importing_red_hat_content#Importing_Red_Hat_Content-Creating_a_Synchronization_Plan

Section Number and Name: 

4.7. Creating a Synchronization Plan

Describe the issue: 

If huge sync plan is created and executed, syncing of repositories will fail after some time

https://bugzilla.redhat.com/show_bug.cgi?id=1635679

Suggestions for improvement: 

Create warning box about possible problems when attempting to sync a lot of repos at once, more reccomend on demand sync and use of smaller batches of repositories in sync plans.

Additional information:

Comment 1 Andrew Dahms 2018-11-07 12:14:09 UTC

Assigning to Sergei for review.

Comment 10 David Davis 2018-11-07 15:18:04 UTC

We really can't specify anything more specific because "large" is highly dependent on the resources of the system--it could be 5 or it could be 50+. I'd recommend perhaps saying that users are encouraged to split up sync plans as much as possible to avoid placing a heavy load on the system.

Comment 12 David Davis 2018-11-07 16:35:34 UTC

I am not sure how useful that would be. There are a lot of variables and some variables might affect how many repositories you can sync at one time. So unless you happen to have the same system specs, it won't be useful.

Comment 16 Pavel Moravec 2018-11-08 12:16:34 UTC

While customer can ask for particular numbers, these rely a lot on too many factors, including (not only):
- number of repos in a product in sync plan
- number of repos/errata/packages-in-errata in a repo
- size of mongo DB (affects search/insert times there)
- I/O performance (physical disk speed, latency, filesystem,..)
- CPU load (candlepin or puppet can generate high load)
- Satellite version (that improves performance a lot, incl. inside pulp)
- number of CPU cores, pulp workers and their proportion

So we might come with some safe low value that is supposed to work either time, and warn "the bigger plan than this you use, the highly probable BZ1635679 happens" ?

Comment 17 Bryan Kearney 2018-11-08 18:09:48 UTC

We would prefer to improve the issues and to provide better tuning guidance. Closing this out as WONTFIX.

Comment 21 David Davis 2018-11-21 17:37:32 UTC

Clearing NEEDINFO.