"When I do a cache flush a red message appears within moment saying it's done" I think the flush is done but that doesn't mean the cache has rebuilt. Kind of like a toilet. When you flush the water goes down fast but it might take a minute for the bowl to fill up again :)
https://docs.google.com/file/d/0B74tdm250wuIbERiSGxNQi1fZ0k Yeah, with our ITSM 8.x prod we only flush at night or in times of emergency. Too much impact otherwise. Sync cache pretty much goes unnoticed unless cache corruption occurs. Jason On Thu, Oct 10, 2013 at 12:34 PM, Susan Palmer <suzanpal...@gmail.com>wrote: > ** > When I do a cache flush a red message appears within moment saying it's > done. But that is false. It seems to take about 15 minutes. I do not > see any status bar. > > I did notice yesterday, since it was my first flush on our new production > server that it affected users in the client and everyone got a timeout, not > just mid-tier users (oooops). We'll call that learning curve! > > I have not tried the sync flush yet, I'll try that tonight. I really > hampers getting things in the system. > > Keep talking about mid-tier, I have a lot to learn! > > > On Thu, Oct 10, 2013 at 2:10 PM, Jason Miller <jason.mil...@gmail.com>wrote: > >> ** >> Actually I think it is more related to Perform Check. I have the Sync >> Cache option in environments where we have Preload on and off. If I >> uncheck Perform Check and save the Sync Cache button turns grey. >> >> >> On Thu, Oct 10, 2013 at 12:04 PM, Arner, Todd <tar...@glhec.org> wrote: >> >>> ** >>> >>> Ah that may be why. We do have preload turned on**** >>> >>> ** ** >>> >>> *From:* Action Request System discussion list(ARSList) [mailto: >>> arslist@ARSLIST.ORG] *On Behalf Of *Jason Miller >>> *Sent:* Thursday, October 10, 2013 2:01 PM >>> >>> *To:* arslist@ARSLIST.ORG >>> *Subject:* Re: Controlling the Flushing action of the Mid-Tier Cache**** >>> >>> ** ** >>> >>> ** **** >>> >>> I have seen it... On the environments where we have Preload turned on.* >>> *** >>> >>> ** ** >>> >>> On Thu, Oct 10, 2013 at 11:47 AM, Joe D'Souza <jdso...@shyle.net> wrote: >>> **** >>> >>> ** **** >>> >>> I have always seen a progress bar after hitting Sync Cache – I’ll take a >>> screen shot next time I have to use it if I remember to and send it to you >>> offline.**** >>> >>> **** >>> >>> Joe**** >>> >>> **** >>> ------------------------------ >>> >>> *From:* Action Request System discussion list(ARSList) [mailto: >>> arslist@ARSLIST.ORG] *On Behalf Of *Jason Miller >>> *Sent:* Thursday, October 10, 2013 2:40 PM**** >>> >>> >>> *To:* arslist@ARSLIST.ORG >>> *Subject:* Re: Controlling the Flushing action of the Mid-Tier Cache**** >>> >>> **** >>> >>> ** **** >>> >>> Ah. So I see a bit of difference in our configurations. Since this is >>> Dev we do not have Preload on. I don't see a status bar when I press Sync >>> Cache. If there were a status bar maybe it would take that long? I just >>> know by the time I check the change in a browser the change is active.** >>> ** >>> >>> **** >>> >>> On Thu, Oct 10, 2013 at 10:59 AM, Joe D'Souza <jdso...@shyle.net> wrote: >>> **** >>> >>> ** **** >>> >>> Wow that’s quick on 8 then. I’m quite certain its not that quick on >>> 7.6.04. Sometimes that progress bar moves a percentage every 3 or 4 or 5 >>> seconds resulting in it getting completed in anywhere between 5 to 10 >>> minutes.**** >>> >>> **** >>> >>> This is probably cause often I’m not the only developer on board and it >>> caches things that others have done too.**** >>> >>> **** >>> >>> My hypotheses without really going through a SQL logging process is that >>> it probably takes quite some time for it to find the objects it needs to >>> cache.**** >>> >>> **** >>> >>> Joe**** >>> >>> **** >>> ------------------------------ >>> >>> *From:* Action Request System discussion list(ARSList) [mailto: >>> arslist@ARSLIST.ORG] *On Behalf Of *Jason Miller >>> *Sent:* Thursday, October 10, 2013 1:25 PM**** >>> >>> >>> *To:* arslist@ARSLIST.ORG >>> *Subject:* Re: Controlling the Flushing action of the Mid-Tier Cache**** >>> >>> **** >>> >>> ** **** >>> >>> I am not arguing the merits of the being able cache a single object or >>> maybe better yet a group of selected objects. My experience with Sync >>> Cache is it is a sub 30 second process on AR 8.0 / MT 8.1. In the last few >>> days I have repeatedly changed a form or AL, let it finish saving (AR >>> server cache mode is a factor here since the server cache needs to reflect >>> the object has changed), hit Sync Cache in IE, CTRL + F5 the form in >>> Firefox. I allow minimal time between those steps.**** >>> >>> **** >>> >>> Jason**** >>> >>> **** >>> >>> On Thu, Oct 10, 2013 at 10:13 AM, Joe D'Souza <jdso...@shyle.net> wrote: >>> **** >>> >>> Actually I have seen that behavior too on 7.6.04 Patch 004 where I >>> noticed >>> the time it takes to flush the entire cache is not that significantly >>> more >>> than using the Sync Cache. >>> >>> My reasoning why that must be as it is, is that it still takes the >>> system a >>> little bit of time to figure out what has changed. Sure the searches for >>> this might be optimized, but it's still a search to identify what needs >>> to >>> be flushed. Even if the search takes only a minute, its still a minute >>> more. >>> >>> I'm guessing it takes a little more than a minute to identify what's >>> changed. On an average I have noticed that a Sync Cache operation takes >>> anywhere around 5 minutes or upwards. >>> >>> The benefit of telling the Mid-Tier what to cache upfront, eliminates the >>> need for the Mid-Tier to search what it needs to cache. >>> >>> PLUS while stuff is still being developed, there's a chance you do not >>> want >>> to cache everything you just developed as you are still R&D'ing it. So >>> as a >>> developer if I want to cache just a single object or a selected list of >>> objects but not the rest that are newly created, I think it would be a >>> huge >>> win if I could have that. >>> >>> I think the Flush Cache and / or the Sync Cache operations are more ideal >>> for end of development cycles or end of day operations. Not middle of the >>> day when you have a team of developers who didn't want the MT cache >>> flushed >>> waiting on it to finish its work. The single object flush idea would be >>> ideal for middle of the day use where you do not impact other users or >>> developers.. >>> >>> Cheers >>> >>> Joe**** >>> >>> >>> -----Original Message----- >>> From: Action Request System discussion list(ARSList)**** >>> >>> [mailto:arslist@ARSLIST.ORG] On Behalf Of Kemes, Lisa A DLA CTR >>> INFORMATION >>> OPERATIONS**** >>> >>> Sent: Thursday, October 10, 2013 12:58 PM >>> To: arslist@ARSLIST.ORG >>> Subject: Re: Controlling the Flushing action of the Mid-Tier Cache >>> >>> We are on 7.6.04 sp2. I'll keep playing with it and see if I can get it >>> to work for me. I'll try clearing the browser cache, etc. >>> >>> Lisa >>> >>> -----Original Message----- >>> From: Action Request System discussion list(ARSList) >>> [mailto:arslist@ARSLIST.ORG] On Behalf Of Jason Miller >>> Sent: Thursday, October 10, 2013 11:41 AM >>> To: arslist@ARSLIST.ORG >>> Subject: Re: Controlling the Flushing action of the Mid-Tier Cache >>> >>> ** >>> What version are you using? I think sync cache has matured a bit since >>> it was first introduced. On 8.0 I am constantly using it to see form >>> and AL updates. >>> >>> One trick is you need to either clear your browser cache or force a >>> cache refresh by using CTRL + F5. Also in the past I have noticed some >>> issues where changed objects being displayed via an inline panel are not >>> easily refreshed. In that case I keep the form itself open in a tab and >>> force the cache refresh on the form itself before trying it inline. >>> >>> Jason >>> >>> >>> On Thu, Oct 10, 2013 at 5:13 AM, Kemes, Lisa A DLA CTR INFORMATION >>> OPERATIONS <lisa.kemes....@dla.mil> wrote: >>> >>> >>> I agree, I tried using the Sync Cache feature and it's good for >>> forms >>> and fields, but not for workflow. I waited and waited for my >>> workflow >>> (active links) to refresh and after about 20 minutes, I just >>> flushed the >>> entire cache and only then did I see my new workflow kick off. >>> >>> -----Original Message----- >>> From: Action Request System discussion list(ARSList) >>> [mailto:arslist@ARSLIST.ORG] On Behalf Of Joe D'Souza >>> Sent: Wednesday, October 09, 2013 6:04 PM >>> To: arslist@ARSLIST.ORG >>> Subject: RFE: Controlling the Flushing action of the Mid-Tier >>> Cache >>> >>> ** >>> >>> Currently the flushing action of the Mid-Tier Cache is limited >>> to either >>> flushing the entire cache (HUGE performance hit) or using the >>> new Sync >>> Cache feature (relatively less of a performance hit). >>> >>> >>> >>> It would be nice if there was another layer of control over what >>> a >>> developer or a Mid-Tier administrator would have liked to flush. >>> And >>> that is the ability to flush an individual object - Just the >>> Form, or >>> just one Active Link, or just one Form Menu, etc. or a selected >>> list of >>> these. >>> >>> >>> >>> The Sync Cache feature already having been built, whose >>> functionality I >>> was a little confused over because it still takes a ton of time >>> to >>> happen, I would not think going one layer deeper would be that >>> much more >>> difficult, and flush just selected objects. >>> >>> >>> >>> This could be implemented both on the Mid-Tier configuration >>> application >>> >>> >>> >>> OR >>> >>> >>> >>> Even better on the Developer Studio itself. The Developer Studio >>> can >>> already store information of various Mid-Tier servers associated >>> with >>> various AR Servers from various environments such as >>> Development, Test, >>> QA, Staging, Production etc. in its Mid-Tier servers >>> Information. It >>> already has the ability to view a form through a browser >>> utility. In >>> version 8 it has even gone a step further to model permissions >>> you want >>> to view the form with. Below is a screen shot of a limited >>> version of >>> this functionality in 7.6.04.. >>> >>> >>> >>> >>> >>> >>> >>> It would be nice to be able to right click on a newly created >>> menu, and >>> perform a Cache on Selected Mid-Tier Server and offer the >>> developer a >>> choice of valid servers associated with that environment that he >>> would >>> like to flush the cache without leaving the Developer Studio and >>> choose >>> single objects or a finite list of objects to flush instead of >>> having >>> the Sync cache flush all the recently developed objects, which >>> may not >>> be what he wants to do. >>> >>> >>> >>> In my opinion this would be less time consuming that even the >>> Sync >>> feature, and will give the developer a better control on what >>> they are >>> ready to publish as finished code to the user. >>> >>> >>> >>> I would like to know what you'll think of this one too. >>> >>> >>> >>> I have yet another couple of ideas I spoke to engineering about >>> that I >>> was asked to post on here so it could be run by the community. >>> >>> >>> >>> Please visit the community area to vote on this one >>> <https://communities.bmc.com/ideas/3588> if you like it at >>> https://communities.bmc.com/ideas/3588. >>> >>> >>> >>> Cheers >>> >>> >>> >>> Joe D'Souza**** >>> >>> >>> _______________________________________________________________________________ >>> UNSUBSCRIBE or access ARSlist Archives at www.arslist.org >>> "Where the Answers Are, and have been for 20 years"**** >>> >>> **** >>> >>> _ARSlist: "Where the Answers Are" and have been for 20 years_ **** >>> >>> _ARSlist: "Where the Answers Are" and have been for 20 years_ **** >>> >>> **** >>> >>> _ARSlist: "Where the Answers Are" and have been for 20 years_ **** >>> >>> _ARSlist: "Where the Answers Are" and have been for 20 years_ **** >>> >>> ** ** >>> >>> _ARSlist: "Where the Answers Are" and have been for 20 years_ **** >>> ------------------------------ >>> The information contained in this communication may be confidential, is >>> intended only for the use of the recipient(s) named above, and may be >>> legally privileged. If the reader of this message is not the intended >>> recipient, you are hereby notified that any dissemination, distribution, or >>> copying of this communication, or any of its contents, is strictly >>> prohibited. If you have received this communication in error, please notify >>> the sender immediately and destroy or delete the original message and any >>> copy of it from your computer system. If you have any questions concerning >>> this message, please contact the sender. >>> _ARSlist: "Where the Answers Are" and have been for 20 years_ >>> >> >> _ARSlist: "Where the Answers Are" and have been for 20 years_ >> > > _ARSlist: "Where the Answers Are" and have been for 20 years_ > _______________________________________________________________________________ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org "Where the Answers Are, and have been for 20 years"
