GC Overhead on ColdFusion 11

Similar Questions

• The repeated CFC function call is much slower than the UDF version

  For a project, we were faced with unexpected behavior of a CFC. Without going to much into the details of the project, it boils down to the following: when we run through a set of records, say 100 and apply a function to this topic, the overhead is 4-5 seconds extra when we call it a component instantiated (only once) compared to when code us as a UDF. Since our application can process thousands of records in 1 past the overload will quickly over 100 seconds compared to the UDF.

  I was wondering if there is no logical explanation for this behavior. From what I understand a function called from a component should not add any notable overload after the component was instantiated.

  Our installation is an installation of Coldfusion 8 Jrun running on Linux. We didn't really find anything on the web documenting this (or we found we have said that there will be no additional charge). We could modify the function call to a call from the UDF, but we prefer to keep in a CFC as there are more things behind the scenes that just formatted data.

  B.

  HA!  Cool!!! I think I solved it something, I feel so proud.

  I run the code of Adam and the first time I ran it I got similar to Biebel3 results.  But it's curious, I got the output of debug it showing that it wasn't show where the time was spent.  He would show almost two full seconds to build the page of CFCS, but the components in the table of debugging only added up to about 50 milliseconds.

  I tired first put a top of the page of CFC, but that has not changed the results.  Because even if you do not display the results, ColdFusion still satisfied the debug output.  I went into the administrator and effectively disabled the debugging feature.

  Alto, I started to have the results displayed by Adam.

  It seems that it is a question where it takes a lot of overhead for ColdFusion compile a record of debugging for each call of CFC in this large loop.

  The good news, this is a pretty easy solution for a production system.

• Could ColdFusion have caused the URLRewrite not work correctly?

  I use ColdFusion 10 with CMS the Paperthin CommonSpot. The hyperlink in each of the search result page is https://devbox.mysite.com/kb/article/? title =test-article and the URLRewrite should show https://devbox.mysite.com/kb/article/test-article on the client instead of the browser. Below is the URL rewriting.

  <rule name="KB-rewrite" enabled="true">
      <match url="^kb/article/([a-zA-Z0-9\-]+)$" />
      <conditions logicalGrouping="MatchAll" trackAllCaptures="false">
      <add input="{QUERY_STRING}" pattern="1" negate="true" />
      </conditions>
      <action type="Rewrite" url="kb/article/?title={R:1}" appendQueryString="false" />
  </rule>
  

  I know that the URL rewrite rule works because if I removed the "? title -"URL, it loads the article very well. The question is, why it shows ? Legislative = in the URL when, without it, it would be still works? What could cause this behavior? Where can I start troubleshooting? Would it make sense to my application using ColdFusion that cause the problem? Suggestions are appreciated.

  Well, I was pretty stupid on this one.  IIS URLRewrite works like she supposed to be so I removed the? title =, the overhead of the page. The problem is in the hyperlink search result, I put? ititle = in the URL so that a user has clicked on the result, it would show the URL to contain the? title = inside. As the URLRewrite works as it should, all I had to do was remove the? title = hyperlink of each result and my problem is solved.

• CFThread GC Overhead limit exceeded

  I have an application that trades virtual items and have a single page that gets all my accounts and for each, we create a thread that initially connects the account and then search and buy items for as long as the session is active.  I must point out at this stage that this is my first experience of using cfthread.

  I have problems with it. Every 30 minutes (if not less) my ColdFusion server stops and I have to restart the service. After restarting the service, I check the logs and there are errors that say "GC Overhead limit exceeded".

  I looked online at length but as far as cfthread is new to me, this is the JAVA virtual machine and how it works. I'm running on CF10 Enterprise Edition and loaded the server monitor and enough surely, I can see the JVM memory usage grow and develop until the limit has been reached (at the time I put it only 2 GB because when I had it set up the memory seemed fills up faster). Even when I select the option run GC in the monitor does not reduce the memory usage a lot, or not at all.

  It's more than likely something to do with my code? For the moment, I have a little less than 50 threads are created, but I added multiple accounts on the application then extra threads that will be needed.

  Here is the code of the page...

  < script >

  / * RELOAD THE PAGE EVERY 65 MINUTES * /.

  {setTimeout (function ()}

  Window.Location.Reload (1);

  (}, 3900000);

  < /script >

  <! - MOVE ACCOUNTS - >

  < cfquery name = "getLogins" datasource = "myDB" >

  SELECT * FROM Logins WHERE active = 1

  < / cfquery >

  <! - LOOP ACCOUNT - >

  < cfloop query = "getLogins" >

  <!-HAVE a SLEEP SO IP is not MARKED FOR SENDING TOO MANY REQUESTS AT the SAME TIME->

  < cfset Sleep (30000) / >

  <!--CREATE THREAD FOR ACCOUNT-->

  < cfthread

  Name = "" #getLogins.AccountName # ""

  action = "run".

  accountName = "#Trim (getLogins.accountName)" # ""

  email = "#Trim (getLogins.email)" # ""

  Password = "(getLogins.Password) #Trim" # ""

  resourceId = "#Trim (getLogins.resourceID) #" >

  <!--> DEFAULT SESSION variables

  < cfset SESSION ["#attributes.accountName #LoggedIn"] = 0 / >

  < cfset SESSION ["#attributes.accountName #LoginAttempts"] = 0 / >

  <!-ACCOUNT NOT CONNECTED AND LESS THAN 8 CONNECTION ATTEMPTS MADE->

  < cfscript >

  While (SESSION [' #attributes.accountName #LoggedIn'] EQ 0 AND ' SESSION [' #attributes.accountName #LoginAttempts] LT 8) {}

  CONNECTION ATTEMPT

  THREAD.logInAccount = Application.cfcs.Login.logInAccount (attributes.email, attributes.password);

  IF LOGIN ATTEMPT

  If {(EQ THREAD.logInAccount 0)}

  INCREASE THE NUMBER OF ATTEMPT

  SESSION [' #Attributes.AccountName #LoginAttempts'] = ' SESSION [' #attributes.accountName #LoginAttempts] + 1;

  }

  ELSE IF RETURNED IS 481 THEN THE ACCOUNT IS LOCKED

  ElseIf (THREAD.logInAccount EQ 481) {}

  SET COUNT OF CONNECTION ATTEMPT TO STOP THE LOOP

  SESSION [' #Attributes.AccountName #LoginAttempts'] = 8;

  UPDATE ACCOUNT TO MARK AS LOCKED

  THREAD.updLogin = Application.cfcs.Login.updLogin (attributes.email);

  }

  }

  < / cfscript >

  <!-If CONNECTED-> ACCOUNT

  < cfif SESSION [' #attributes.accountName #LoggedIn'] EQ 1 >

  <! - SET ID FOR RESEARCH - >

  < cfset THREAD.definitionID = attributes.resourceID - 1610612736 / >

  <!-CONNECTED-> ACCOUNT

  < cfloop condition = "SESSION [' #attributes.accountName #LoggedIn'] is equal to 1" >

  <! - GET LATEST more LOW BUY it NOW PRICE - >

  < name = "Cfquery THREAD.getMinBIN" datasource = "WAS" cachedWithin ="#CreateTimeSpan (0,0,1,0) #" > "

  SELECT TOP 1 * FROM v_FUT14BINPrices WHERE resourceID = #attributes.resourceId # ORDER BY DESC lastUpdated

  < / cfquery >

  <!--INCLUDE FILE THAT CALCULATES the PURCHASE AND SELLING PRICE-->

  < cfinclude template = "sellingPrices.cfm" / > "

  <! - IF the PRICE of the tender has BEEN SET - >

  < StructKeyExists (THREAD, "biddingPrice") cfif >

  <!--> SEARCH MAKE, REQUEST APPLICATION

  < cfset THREAD.requestStart = GetTickCount() / >

  < cfset THREAD.search = Application.cfcs.Search.dosearchOld(attributes.resourceId,THREAD.biddingPrice,0) / >

  < cfset THREAD.requestDuration = GetTickCount() - THREAD.requestStart / >

  <! - IF the SEARCH CONTAINS the CONTENTS OF THE FILES - >

  < StructKeyExists (THREAD.search, "FileContent") cfif >

  <! - STATE the NUMBER OF RESULTS VARIABLE - >

  < cfset THREAD.numResults = 0 / >

  <!-If RETURNED-> JSON

  < IsJSON (THREAD.search.FileContent) cfif >

  <! - DESERIALIZE JSON - >

  < cfset THREAD.searchResults = DeserializeJSON (THREAD.search.FileContent) / >

  <! - IF PLAYER SEARCH RETURNS STRUCT AUCTIONINFO - >

  < StructKeyExists (THREAD.searchResults, "auctionInfo") cfif >

  <! - SET NUMBER OF CARDS RETURNED BY SEARCH - >

  < cfset THREAD.numResults = ArrayLen (THREAD.searchResults.auctionInfo) / >

  < cfset THREAD.statusCode = "Successful" / >

  < cfif THREAD.numResults EQ 0 >

  < cfset THREAD.statusCode = "Successful - no. Results" / >

  < / cfif >

  <!- OTHERWISE, if the CODE of ERROR RETURNED-->

  < cfelseif StructKeyExists (THREAD.searchResults, "code") >

  < cfset THREAD.statusCode = THREAD.searchResults.code / >

  <!-IF CODE 401 THEN SESSION has EXPIRED->

  < Cfif THREAD.statusCode EQ 401 >

  <!-SESSION as it is DISCONNECTED from the VALUE AND TRY SESSION REFRESH->

  < cfset SESSION [' #attributes.accountName #LoggedIn'] = 0 / >

  < cfset THREAD.logInAccount = Application.cfcs.Login.logInAccount (attributes.email, attributes.password) / >

  < / cfif >

  <! - ANOTHER THING HAPPENED - >

  < cfelse >

  < cfset THREAD.statusCode = "Something Else" - & THREAD.searchResults.code / >

  < / cfif >

  <!-If the RESULTS RETURNED--->

  < GT 0 cfif THREAD.numResults >

  <!-LOOP AROUND OF RESULTS AND CHECK if the CRITERIA of PURCHASE of GAME - >

  < cfloop index = "i" = "1" to = "#THREAD.numResults #" >

  <!-* SAFETY CHECK * - make sure THAT the ID OF the CURRENT MAP IS the SAME AS for a RESEARCH - >

  < cfif THREAD.searchResults.auctionInfo [i].itemData.resourceID EQ attributes.resourceId AND THREAD.getMinBIN.resourceID EQ attributes.resourceId >

  <! - ENSURE BIN PRICE TOGETHER AND IS LESS than the SET PURCHASE PRICE - >

  < cfif THREAD.searchResults.auctionInfo [i] .buyNowPrice GT AND THREAD.searchResults.auctionInfo [i] .buyNowPrice LTE THREAD.biddingPrice 0 >

  <!-GAME-> END of the auction TIME

  < cfset THREAD.timeLeft = THREAD.searchResults.auctionInfo [i] expires / >

  < cfset THREAD.auctionEnds = DateAdd ("s", THREAD.timeLeft, Now ()) / >

  <! - BUY CARD - >

  < cfset THREAD.buyCard = Application.cfcs.Bid.doBIN (THREAD.searchResults.auctionInfo [i] .tradeID, THREAD.searchResul ts.auctionInfo [i] .buyNowPrice, THREAD.searchResults.auctionInfo [i] .startingBid, THREAD.searc hResults.auctionInfo [i].itemData.ID, THREAD.searchResults.auctionInfo [i].itemData.resourceI D, THREAD.startPrice, THREAD.binPrice, THREAD.lowestBIN, THREAD.searchResults.auctionInfo [i] .i temData.discardValue, THREAD.auctionEnds, THREAD.requestStart, THREAD.requestDuration) / >

  < / cfif >

  < / cfif >

  < / cfloop >

  < / cfif >

  < cfelse >

  < cfset THREAD.statusCode = THREAD.search.FileContent / >

  < / cfif >

  < cfset THREAD.sleepDuration = 1000 - THREAD.requestDuration / >

  < cfif THREAD.sleepDuration GT 0 > < cfset Sleep (THREAD.sleepDuration) / > < / cfif >

  < / cfif >

  <!--> INSERT RECORD SEARCH

  < cfset THREAD.insSearchRecord = Application.cfcs.Search.insSearchRecord (THREAD.definitionID, THREAD.statusCode, THREAD.requ, estDuration, THREAD.numResults, THREAD.biddingPrice) / >

  < / cfif >

  < / cfloop >

  < / cfif >

  < / cfthread >

  < / cfloop >

  I would have thought that memory would have stayed around the use as well as each loop performs the same set of actions so once the loop has returned at the beginning then I thought the previous loop would have been removed from memory (free space) and then the same actions would be made so the same total memory use then upwards , but it almost seems as if each loop is kept in memory and it is therefore more and more.

  Could someone please help me and offer some advice on how I could fix this problem? If you need more info, let me know

  Thanks in advance

  

  JVM looks like this thanks gcviewer:

  The diagram does not show the PermGen details. The tail of the log says:

  Total PSPermGen 1048576K, used space K 80172 object 1048576K, 7% used.

  So that you have something to do with the CFM code, for the time being to maintain the system, you may do better to make a few adjustments of the JAVA virtual machine. Remember, that this is probably not a fix for the problem in his together more try this to see if the system remains standing while you continue to work on other issues CFM.

  I'm waiting for the JVM details are like that:

  Minimum size: 1024
  Maximum segment size: 2048
  -XX: MaxPermSize = 1024 m

  The entire system is:

  CF10 on one machine Windows 7 x 64 with an Intel core i3-2100 processor and 8 GB of RAM.

  Not seeing is not a lot of objects is kept in PermGen you can do which is smaller. Bunch of double size and set a value for the new part of the bunch. FOR EXAMPLE:

  Minimum size: 2048
  Maximum segment size: 4096
  -XX: MaxPermSize = 324 m
  -Xmn256m

  Or that way if you prefer in CFadmin:

  -serveur - Xmn256m - XX : MaxPermSize = 324 m - XX : PermSize = 192 m - XX : + UseParallelGC - Xbatch-Dcoldfusion.home={application.home}-Dcoldfusion.rootDir={application.home}-Dcoldfusion.libPath={application.home}/lib-Dorg.apache.coyote.USE_CUSTOM_STATUS_MSG_IN_HEADER=true-Dcoldfusion.jsafe.defaultalgo=FIPS186Random - XX : + PrintGCDetails - XX : + PrintGCTimeStamps - XX : + PrintHeapAtGC-verbose : gc-Xloggc:cfjvmGC.log

  Or it by editing the JVM. CONFIG:

  -serveur - Xms2048m-Xmx4096m-Xmn256m - XX : MaxPermSize = 324 m - XX : PermSize = 192 m - XX : + UseParallelGC - Xbatch-Dcoldfusion.home={application.home}-Dcoldfusion.rootDir={application.home}-Dcoldfusion.libPath={application.home}/lib-Dorg.apache.coyote.USE_CUSTOM_STATUS_MSG_IN_HEADER=true-Dcoldfusion.jsafe.defaultalgo=FIPS186Random - XX : + PrintGCDetails - XX : + PrintGCTimeStamps - XX : + PrintHeapAtGC-verbose : gc-Xloggc:cfjvmGC.log

  Because now, it would be wise to put the JAVA virtual machine to run the full garbage collection every 10 minutes, but I post I'm undecided if will show even more thoughts on that if she seems more like an idea of good dressing for me. I think stay with UseParallelGC because it tends to keep the evacuated heap rather than change the garbage collector to something that tends to maintain the objects in memory.

  Java is 7 so no 6 EOL tho just to say that 7u15 is old with current 7u67. Java 8 is also communicated however not substantiate the claim of Adobe with execution CF10 on Java 8. Given that you do not use a new garbage collection as G1GC algorithim (and for now, I do not recommend do you) I think stay with 7u15 for now.

  HTH, Carl.

• Mac Book to overhead.

  How can I connect my Mac Book to an overhead projector?

  You will need the appropriate adapter and cable.

• When I connect my iPod to iTunes it says the storage overhead but it says on the iPod 8 GB available

  Please help me with this

  Looks like you have iTunes configured to synchronize automatically when a device is connected and you have iTunes configured so that one tries to synchronize media more to the iPod as the storage space available on the iPod.

  Check your synchronization settings:

  iTunes: synchronization of multimedia content for iPod and iOS devices

  https://support.Apple.com/en-us/HT201253

  Try the synchronization using the manual method

  Managing content manually on iPhone, iPad and iPod

  https://support.Apple.com/en-us/HT201593

  The error message really says the 'storage overhead '?

• Overhead of the graphic use of Wfm as a Subvi output terminal?

  -My basic question:
  If the front is closed, is there a difference between a graph of Wfm and Wfm indicator as output terminal?

  -The question in context:
  I have a no reentrante Subvi, which will take place in my program and an output waveform.  Sometimes, I want to open the front of the Subvi and adjust its internal settings displaying its output on a graph of Wfm.

  The minimum solution would appear to make the graph of Wfm of the Subvi double panel as the output terminal, as opposed to the output through an indicator of Wfm.  But who would increase the resources used during normal operation of the Subvi?

  Thanks in advance and have a nice day.

  -Brad

  There should be no overhead. The chart would update only if the front is open. Most of the time, there are not yet signs before the subVIs in memory.

  Mike...

• TCP / HTTP overhead

  I apologize if this isn't the right place to ask this question... I'm trying to understand overhead with tcp and HTTP response that I see in the capture of packages (wireshark) that I join you the topic.

  My understanding is:

  I can calculate the portion of TCP data by subtracting the ip/tcp headers of the total length field in the IP header. My confusion is when watching the payload data tcp and then seeing the overhead that is specified in the body of message/response HTTP header.  In my view, it has 1448 bytes which corresponds to the part of the packet tcp data.

  However, the HTTP response header is 347 bytes and the Content-Length entity 3867 bytes message body. I am trying to wrap my head on how to determine the correct load for this specific package. Normally, it's very simple, but it's the rsponse HTTP header which throws me off the power.

  Can someone break down and help me understand how I can have 1448 for TCP data but larger values for the HTTP part?

  Because the HTTP message is fragmented.

  You can see the somdwhere message back in the next Wireshark screen.

• IPSec overhead

  Hello

  Can someone tell me what kind of overhead 3DES (ESP) puts on an IP packet?

  Memory, theres 50-73 extra bytes and I remember that it must be in increments of 8 but I don't understand the huge variation (50-73).

  Can I accurately calculate the load 3DES on IP?

  Concerning

  Scott

  The load depends on the set of transformation.

  for esp-3des esp-md5-hmac, esp - the hmac-md5-esp, esp-3des esp-sha-hmac, esp - of the overhed esp-sha-hmac will be 50-57bytes

• GET overhead VPN

  Hi all

  We are looking for overhead represented due to GET VPN. Is there a table of comparison or the value.

  Thank you

  Concerning

  Anantha Subramanian Natarajan

  Anantha,

  As mentioned by Lloyd, GETVPN the new IP header is a copy of the original IP header. So, who's going to be 20 bytes (without options). Please keep in mind that the size of the package may vary depending on the options of encryption and authentication as AES, SHA, etc.. Basically, around 52 to 56 bytes. Thus, with the new IP header, looking at 72 to 76 bytes.

  I will quote the ESP RFC 4303 for more details.

  I have not seen a document of specific performance GET VPN on cisco.com. But, since the Original IP header is copied and placed in front of the ESP instead of a new IP as the traditional IPSEC header, I don't think he'll be a lot of difference in the performance of encryption between traditional and GET VPN.

  I hope it helps.

  Kind regards

  Arul

• Upgrade from ColdFusion 2016: can't access folders /api/

  Hello CF community.

  I recently upgraded from ColdFusion to ColdFusion 2016 10 and I noticed that I can access is no longer in all pages who live in a folder named/api /.

  For example: www.mysite.com/api/index.cfm now returns a 500 error "application could not be found.  See the following screenshot:

  http://i.imgur.com/axFVekv.PNG

  Note: when I installed ColdFusion 2016 I did not choose to install the service manager of the API or do I need.

  Thanks for your help!

  This is because there is a servlet-mapping/api cf /.

  You can remove the mapping by going to cfinstall/cfusion/wwwroot/WEB-INF/web.xml.  You can search the api servlet-mapping and comment him.

  HTH,

  -Dave

• Error installing Coldfusion 10 Update 21

  Hello

  I'm having trouble installing the latest patch on my developer edition 21 CF. I have not had problems before with the installation of patches.

  I want to install using the Windows command line, as the CF Admin has never worked method (impossible to download patches because of our network problems)

  3 fatal errors in the log file:

  1. ERROR - Could not move the C:\{CFPATH}\lib\updates\{filename}.details file to the location of backup C:\{CFPATH}\hf-updates\hf-10-00021\backup\lib\updates\{filename}.details
  2. ERROR - Could not move the C:\{CFPATH}\lib\updates\{filename}.notes file to the location of backup C:\{CFPATH}\hf-updates\hf-10-00021\backup\lib\updates\{filename}.notes
  3. ERROR - Could not move the C:\{CFPATH}\lib\updates\{filename}.hotfix_021.jar file to the location of backup C:\{CFPATH}\hf-updates\hf-10-00021\backup\lib\updates\{filename}.hotfix_021.jar

  I am running the command-line in Windows as an administrator. So I don't think it's a problem of permission. All the more that the other files are written in these places during the installation (C:\{CFPATH}\hf-updates\hf-10-00021\backup\...).

  OS is Windows 7

  During the attempt to install the fix, I see the files *.details and *.notes writing in the C:\{CFPATH}\lib\updates folder. They are deleted at the end of the process.

  No idea why?

  Thank you.

  Had a similar problem with update 20. Here area few things Adobe told me to do.

  1. Make sure that you open the command prompt with run as administrator. You have the right permissions.
  2. They asked me to run the update with the version of JAVA that the ColdFusion server is / was using. Seems that the version of Java that I had for the operating system was not the same version as the one referenced in ColdFusion server. So, I found myself by running the command as: "C:\Program Files\Java\jre1.8.0_101\bin\java.exe" - jar ".jar JAR file c:\{name}"

• ColdFusion 11 Enterprise vs standard

  Hi all

  We plan to upgrade our Enterprise version 10, because it would not connect to our database company Oracle 12 c.

  CF 11 Standard giving us that CF 10 Enterprise does not work and it would be able to connect with Oracle 12 c enterprise edition?

  We have installed CF on several servers. That is the key to the standard (installation on one server only) vs company (can install on several servers) and CF 11 would connect with 12 c Oracle enterprise server.

  Thank you

  Hello

  Since this is a 3rd pilot, you can't find falling down. Please follow the instructions below.

  1. download the driver think Oracle's Web site.

  2. you can copy the driver \ColdFusion\cfusion\lib or ColdFusion\cfusion\wwwroot\WEB-INF\lib

  3. restart the ColdFusion service.

  4. connect to CFadmin > Datasources > add data source name and the menu drop down select 'other '.

  5 URLS to connect to Oracle are

  a. If you have configured the SID, follow the below URL

  JDBC: oracle:thin:@//{HOSTNAME}:{PORT}/SID

  b. If you have configured the service name, follow the below URL

  JDBC:Oracle:thin:@//{hostname}:{port}; ServiceName = {ServiceName}

  6 class driver: oracle.jdbc.driver.OracleDriver

  7. Enter the user name and password and validate.

  I hope this helps.

  Thank you

  Priyank

• 2016 ColdFusion and IIS 8.5

  Hello

  I can't fix the below error while configuring IIS 8.5 for my ColdFusion 2016, please fix for the issue.

  Thank you in advance .

  

  Hello Aneesh,

  Could you please recreate the connector using the following procedure:

  1) go to the ColdFusion2016\cfusion\runtime\bin directory and run wsconfig as 'administrator '.

  (2) remove the connector for this site

  (3) add the connector

• ColdFusion Server Monitor error

  Hi all

  OS / Redhat Linux 7

  CF Version 11,0,07,296330

  Tomcat 7.0.64.0 version

  Java version 1.8.0_25

  I have this strange problem that I hope is an easy solution.  Already when I ran CF9 I was able to enter the server without problem monitor.  In June, we have deployed a new server and installed CF11.  Since then, I was not able to get in the server monitor, so I guess that the problem is related to the way that configure ColdFusion server.

  When I click on Start Server Monitor in the CF administrator, he launches a new window, but displays the following error message:

  

  I don't know if he just needs in bad the monitor server, or the issue is due to the account used to set up see  ColdFusion is running as user "nobody"... which I'm sure that isn't the right way to have this set up.  For login information, I tried using the administrator account that connect to the Admin of CF page with, but did not have luck.

  Any help for that to be resolved would be greatly appreciated

  Thank you!

  We ended up finding it works. It turns out that the port 5500 has been blocked at the firewall level... ugh.

  Thank you!