This is a problem I've dealt with randomly for many years now.
At my new company, our Full sandbox has what I like to call a "caching" issue. I know no other way to describe it.
Every morning, I am the first at work and the first in our Sandbox (frequently the only one all day).
When simply viewing, or inserting or editing, a record into the Full sandbox for the FIRST time for EACH object, it takes 15-60 seconds for the appearahce, or save, to happen (assuming no Apex or other intervention - then it could be up to 2 minutes).
Even visiting the Apex dev pgaes takes 15-45 seconds the first time "per day" (though really it's "per several hours").
AFTER that first touch inside any given object, that object THEN works JUST AS IT SHOULD, speed-wise.... until you stop touching it for several hours. Then it happens again. This equates to "caching" putting it simplistically. "Priming the pump" as some of my users put it.
Anyone experience the same thing? This is the second Sandbox I've had this happen in officially. Have heard it from another Admin I used to work with also.
Anyone know the right "keywords" to get SF to NOT spend weeks in a Ticket discussing this, deflecting it to a code issue, plugin issue, user issue, etc, so that they can flip the proverbial switch to FIX IT?
The person in charge of our Org prior to me (a contractor) spent 3 months in an SF ticket, chasing down ghosts involving:
- Rootstock
- Financiah Force
- SF Country/State Picklists
...only to have (me) back at square one.
UPDATE 2017-01-16: Sent log of access to SF Support as they state they will check my audit trail and see why the delays are happening. Attaching log here as a reference with Org-identifying info redacted.
UPDATE 2017-03-13:
- Salesforce has isolated the problem to their "precompile" feature (which is not firing when it should to keep Apex fresh all over the Org - not tied to any managed package - it happens if you visit the Apex Classes page, it happens when visiting a record, creating a record, etc, first time every 'few hours'
- I tested in the Production org (where Sandbox originated from) and it has teh SAME PROBLEM (and is just as slow, which is bizarre considering Production typically has "more resources" and you'd at least expect it to do it "faster" - it does not)
- Salesforce has abandoned the situation after almost six months across two cases.