'idp.service.metadata.checkInterval' on specific time

[Ryan Larscheidt]

We found that -XX:G1ReservePercent=25 (with a 4GB heap) was effective in preventing the "to-space exhausted" events when reloading the InCommon aggregate.

I added a comment with our settings to the IdP Heap Management wiki page: https://wiki.shibboleth.net/confluence/display/IDP30/IdP+Heap+Management?focusedCommentId=55804870#comment-55804870

Ryan
________________________________________
From: users <users-bounces at shibboleth.net> on behalf of Zico <mailzico at gmail.com>
Sent: Tuesday, March 19, 2019 09:40
To: Shib Users
Subject: Re: 'idp.service.metadata.checkInterval' on specific time

A quick question... anyone using initial memory allocation ( Xmx ) and MaxMetaspaceSize?
I am planning to use "-Xms3072m -Xmx3072m -XX:MaxMetaspaceSize=256m" in production system.

On Mon, Mar 18, 2019 at 12:08 PM Cantor, Scott <cantor.2 at osu.edu<mailto:cantor.2 at osu.edu>> wrote:
On 3/18/19, 12:52 PM, "users on behalf of Jim Fox" <users-bounces at shibboleth.net<mailto:users-bounces at shibboleth.net> on behalf of fox at washington.edu<mailto:fox at washington.edu>> wrote:

> We had a problem with InCOmmon metadata reloads causing memory to fill.
> It turned out for us that we were triggering manual metadata reloads too
> often.

I'm on a 3 hour interval, but of course it mostly just no-ops due to it being unchanged outside of each afternoon. It should do that if you manually reload it also (via the actual reload trigger, vs. actually forcibly updating a local copy on disk being read).

-- Scott


--
For Consortium Member technical support, see https://wiki.shibboleth.net/confluence/x/coFAAg
To unsubscribe from this list send an email to users-unsubscribe at shibboleth.net<mailto:users-unsubscribe at shibboleth.net>


--
Best,
Zico