Directory Watcher is Skipping files

I've noticed the directory watcher is skipping files when a
number of files appear in the directory in less than a minute. The
first one or 2 are successfully processed, but the next one(s) are
skipped.
The timestamps on the files are like:-
10:28:32 done
10:29:06 done
10:29:14 done
10:29:17 skipped
What is causing this and what can I do to fix it?
Chris

In article <fc5jdn$pjr$[email protected]>
"105"<[email protected]> wrote:
> I've noticed the directory watcher is skipping files
when a number
> of files appear in the directory in less than a minute.
The first one
> or 2 are successfully processed, but the next one(s) are
skipped.
Well, the directory watcher is supplied as an *example* and
is not
really intended to be production-quality code. The source is
supplied
so you are free to modify it yourself as needed.
Sean A Corfield -- (904) 302-SEAN
An Architect's View --
http://corfield.org/
cf.Objective() 2008
Early May, near Minneapolis, MN
The world's ONLY enterprise engineering conference for
ColdFusion
http://cfobjective.com/

Similar Messages

CF9 Directory Watcher randomly skips files

I'm using the directory watcher gateway to log changed files as they are pushed (FTP'd) to a development server. The gateway occassionally misses random files. I've seen some older post where this was a problem and was looking to see if the problem still exists on CF9. The onAdd/onUpdate methods simply write the file details to a SQL database. There is very little logic in the .cfc.
If you FTP 10 files to the server it will tell you 8 changed. If you push the same 10 again it will in some cases it will reconcile the missing files. I have the intervalconfigured for 60 seconds.

My problem was not "randomly skipping of files". In the ColdFusion administrator under Event Gateways/Settings a parameter for "Maximum number of events to queue" is configured to 10 by default. If you FTP more than 10 files in a timeframe under your configured run interval you will lose events. My directory watcher was running every 60 seconds and it was common to FTP > 10 files in 60 seconds. My solution was to increase the "Maximum number of events to queue" and decrease my run interval from 60 to 15. The combination of these two configuration changes has fixed my missing files problem.

Directory watch on another server

OK, we have a weird situation. I am in an environment where
we have ColdFusion MX 7 on our development server but CF 5 on our
production servers. I have developed a directory watcher solution
to automate one of my processes but it will be at least 6 months
before we have converted our nearly 100 apps to MX. Can I use my
development server running MX to watch a directory on another
server? Perhaps using a UNC path in the config file?

Hi,
ups - forgotten this one. The running Oracle process does not have access privileges for mapped network drives by default. So you need to
setup the Oracle service in a manner that the process can "see" the mapped network drives .. That is windows-specific ...
Maybe these articles help ...
http://asktom.oracle.com/pls/asktom/f?p=100:11:4352309546273981::::P11_QUESTION_ID:241814624807
http://www.dbforums.com/oracle/983402-utl_file-network-copy-problem.html
Does this help ..?
Best regards
-Carsten

Backup skips files

Hi
I have a strange error with my Time Machine backup.
The backup is skipping newly created files.
The only way I can get the backup to include the new files is to go in to the very top level "Macintosh HD" or other directory and change a file so that the #changed" date is changed at the top level.
Sometimes that does not help if files are very deep in the structure i suspect ...
I did reinstalled the system and started over the time machine backup. Things was going fine for a couple of days. But soon started to skip files again.
Sometimes a restart does help. And sometimes not.
I have also tried "safe reboot" and that helped.
But the problems always comes back after some time.
What can cause this kind of trouble?
Could it be that the user account is corrupted?
I have a test account that seems to be backed up ok. But I have not tested for very long.
Could it be the actual hard drive? I have one of the Imacs with Seagate drives that is part of exchange program.
I don't get any error messages or anything.
Any help would be apreciated.
Regards
olle

If you use the third-party application "Dropbox," you may see files in the Finder that are not present locally, but are actually on a network server. Those files won't be backed up and won't appear in Time Machine, even though they show in a Finder window. If the files in question are in a Dropbox folder, disable the software or refer to its developer for support.
Continue only if you've ruled out Dropbox or any similar product as a possible cause of the issue.
Time Machine uses the modification times of folders to determine whether their contents need to be scanned for changes made since the time of the last snapshot. For some reason, that criterion occasionally fails, and TM doesn’t scan folders with contents that have changed. You may be able to get backups to resume by updating the modification times of all your folders.
Important: This procedure will change the modification times of all folders on the startup volume. It won’t affect anything that’s not a folder, and it won’t change your existing backups at all. If you have a reason to want to keep the modification times of folders as they are, STOP.
Before proceeding, you must have a current backup of all data. Since your existing Time Machine backups may not be current, you’ll have to back up in some other way; for example, by starting a new TM backup set on another external drive. It’s a good idea to do this anyway, as you always need more than one backup, even if everything is working perfectly.
Triple-click anywhere in the line below to select it:
sudo find / -xdev -type d -exec touch {} \;
Copy the selected text to the Clipboard (command-C).
Launch the Terminal application in any of the following ways:
☞ Enter the first few letters of its name into a Spotlight search. Select it in the results (it should be at the top.)
☞ In the Finder, select Go ▹ Utilities from the menu bar, or press the key combination shift-command-U. The application is in the folder that opens.
☞ Open LaunchPad. Click Utilities, then Terminal in the icon grid.
Paste into the Terminal window (command-V). You'll be prompted for your login password. Nothing will be displayed when you type it. If you don’t have a login password, you’ll need to set one before you can run the command. You may get a one-time warning not to screw up. Confirm. You don't need to post the warning. If you see a message that your username "is not in the sudoers file," then you're not logged in as an administrator.
The command may take a noticeable amount of time to run. Wait for a new line ending in a dollar sign (“$”) to appear. You can then quit Terminal.
The next snapshot will take longer than usual, because all folders will be scanned for changes. When it’s complete, check the results to see how much data has been backed up.
This procedure will only affect the backups of your startup volume. If you back up other source volumes with Time Machine and are having the same problem with those, you’ll need to take additional steps.

Cp: skipping file x' as it was replaced while being copied

Sorry if this has been asked before but:
While copying files from another server (cifs) I get the following error:
cp: skipping file `xxxxxxx', as it was replaced while being copied
dmesg says:
CIFS VFS: Autodisabling the use of server inode numbers on \\192.168.xx.xx\xxxx. This server doesn't seem to support them properly. Hardlinks will not be recognized on this mount. Consider mounting with the "noserverino" option to silence this message.
I am unable to copy any files across and if i manage to they are corrupted.
I have mounted the server as such without luck
mount -t cifs //192.168.xx.xx/directory /mnt/floppy/ -o username=xxxxxxxx,password=xxxxxxxxxx,noserverino
Any ideas?
Regards,
Mike.
Last edited by DrMikeDuke (2010-08-04 18:58:02)

An "other" that large usually indicates that the "other" includes corrupted files. Usually restoring from backup eliminated the corrupted files. However, sometimes restoring to factory settings/new iPod is required.
To restore from backup see:
iOS: How to back up
To restore to factory settings/new iPod see:
iTunes: Restoring iOS software

Time Machine skipping files and folders

Running OS X 10.10.2 on a Early 2009 Mac Pro. System installed on a 256 GB solid state drive. 10 GB of memory.
I'm having a problem with Time Machine where it's not backing up some of my files and folders. There doesn't seem to be much rhyme or reason to it, but it includes very, very important things like the parent directory for projects I'm working on, arbitrary hash directories in some git repositories, an assortment of emails, Messages attachments, random preferences files, and somehow the entire contents of my Pictures folder. And probably more -- I haven't checked everything.
I've recently forced Time Machine to do a full traversal when it made a recent backup (and confirmed that it was a full traversal by looking at backupd's output in Console.app), but these files are still missing.
Does anyone have any suggestions for what I can do to fix Time Machine? I just want it to do a full backup, not skipping anything.
As a last ditch effort I could blow away my Time Machine drive and start from scratch, but I really, *really* don't want to do that because it contains several months worth of backups, and I occasionally do need to go back and get older versions of files.

It is a nasty bug of OSX 10.9 and 10.10 . Some workaround is described here: http://apple.stackexchange.com/questions/170219/time-machine-is-skipping-files-a nd-folders-how-to-fix/170336#170336 by Pete.

Is the directory watcher multi-threaded once it calls the cfc?

the docs say there's a single thread that watches the directory. But say for example it finds 3 files and calls the cfc. At that point do all 3 files run one at a time on a single thread in the order they came in, or do they run at the same time on multi-threads?

I think I understood your questions. There were more than one, some implicit. Here are my answers:
ColdFusion runs the directory-watcher in a single thread.
During the execution of the same thread, ColdFusion detects whether of not a file has, or any number of files have, been added, deleted or changed.
When ColdFusion calls the CFC, it does so in that same single thread. My guess is that any file add-events, delete-events and change-events are queued and handled one after the other.

Remote Directory Watcher

Hi!
I have to implement a directory watcher, the directory being on a different remote server.
I found [this solution|http://twit88.com/blog/2007/10/02/develop-a-java-file-watcher/] (http://twit88.com/blog/2007/10/02/develop-a-java-file-watcher/) and was wondering if the directory declaration using the java.io.File class would accept, and if so how, a remote pathname?
If not, would there be another solution?
Thx.

If you're on a Windows machine and you're trying to access shared folders, then using the usual [UNC syntac|http://en.wikipedia.org/wiki/Path_(computing)#Uniform_Naming_Convention] for Windows will work with File objects i.e. you can create a new File() for \\myServerName\mySharedFolder
But keep in mind that this works because Windows handles the access to that folder, not Java. So if you're prompted for a username/password when you try to access the share in Windows Explorer, it'll most probably not work from within your Java app. And like I mentioned earlier, this will not work for non-Windows systems. You could give jCIFS a shot if you want to access Samba shares (this is what normal Windows shares also use) from any platform.
On a related note, keep in mind that Windows maintains a separate creation timestamp for files. But Java only lets you read the lastModified timestamp. Consider that I create a file at 4 PM (creation time) and then edit it at 5 PM (last modified). If I then make a copy of this file at 6 PM, this copy will have it's creation time set to 6 PM, but it'll have its last modified time set to 5 PM, the modified time of the original. Makes sense when you look it one way, but it could trip you up (like it did me :) ). If you must have the creation time instead of the last modified time, you can check out FileTimes by Roedy Green (of Canadian Mind Products fame) which uses JNI to get it.

Directory watcher Flase event

Hi,
I am having a problem with directory watcher, environment is :
Windows 2003
Coldfusion 8.0.1
Current scenario :
My directory watcher is watching a shared folder (A) on network, and copies files from that folder to a different share folder (B) on network (Both shared folder are in different boxes)
Problem :
Sometime my onAdd fires at odd times and it starts copying all the files available on A folder to B. And when I check logs, that time no files were added in A folder.
Problem is onAdd event fires any time and I dont know why its happening.
Please Help.
Some more info:
I am not moving files from A to B, just copying, so old files remains in same folder, there are around 3000-4000 files as of now in A folder, and every day 10-12 files gets copy in A.

In the onAdd method, get Coldfusion to
read the file and then
append it.
<cffunction name="onAdd" output="no">
<cfargument name="CFEvent" type="struct"
required="yes">
<cfset data=CFEvent.data>
<cfset addedFile=data.filename>
<cffile action = "read" file =
"{absolute_path_to_watcher_dir}\#addedFile#" variable =
"textToAdd">
<cffile action = "append" file =
"{absolute_path_to_some_other_dir}\someFile.txt" attributes =
"normal" output = "#textToAdd#">
</cffunction>

How can I convert the volume directory into a single file installer?

How can I convert the volume directory into a single file installer? I would like to hide all the miscillaneous files that I don't care for and be able to have the installer double click a single file and have it automatically install.

On the second prompt screen when prompted 'What kind of self-extracting Zip file file do you want to make?'
Are you choosing the second option (self-extracting Zip file for software installation)?
I have a word file that I created to help me remember - is there anyway to email it to you?

Illustrator needs a Relative mode switch or ability to set a Working directory for each AI file!

Hi guys,
I thought I would post in here because I find I'm wanting this feature more and more over the past year.
This can apply to most of Adobes application but it would be great to see some file management functionality built into the applications themselves. This could then be further expanded in Bridge/Cue to incorporate more applications into the working project but at the very least the application itself the ability to control its own root existence.
On so many occasions I require Illustrators file linking feature to be relative and NOT absolute which it currently is. The only relative capability it has is if ALL the linked assets are in the same local directory as the AI file itself and this as everyone knows breaks the first rule of proper file management.
Many of the projects I work on uses many AI files with sometimes over 50 linked in assets in a variety of different formats from a variety of different clients (locations). I cannot easily move this project folder around without the painful task of linking all the data again and this creates a massive issue.
Ideally an approach similar to Autodesks Maya and Max where you can set project directories. This means all the Save As and other operations will default to that working directory for that file. All linked data will then use the projects working directory as its root and allow the project to move from location to location and not be affected.
This would Im sure benefit many other designers who work on projects with large amounts of data that may not live always in the same spot. For us we work with so many different people during a production that mobility is a must and at the moment we are finding it very difficult.
As always, thanks for reading.
Cheers
Nick.

Hello,
First command
So what does the wiki meen by: "DEVPATH sets the physical device. You can determine this by executing the command
readlink -f /sys/class/hwmon/hwmon-device/device | sed -e 's/^\/sys\///'"
Physical device of what, and what output am I suppose to get: Nothing, a list with output on where the symlinks lead, something else... In my case I got no output, if that is right, I do not know - since I do not understand what the command do.
Seccond comand
DEVNAME: Sets the name of the device. Try:
sed -e 's/[[:space:]=]/_/g' /sys/class/hwmon/hwmon-device/device/name
Does it meen like this - If I during my pwmconfig used hwmon1 wich was coretemp, and hwmon2 wich was nct6775, I should do:
sed -e 's/[[:space:]=]/_/g' /sys/class/hwmon/hwmon1-device/device/name coretemp
sed -e 's/[[:space:]=]/_/g' /sys/class/hwmon/hwmon2-device/device/name nct6775
And that will direct every occurrence of hwmon1 and hwmon2 to the correct sensor chip?
Regards
Martin
Last edited by onslow77 (2015-01-23 21:46:04)

How to Save XML file in the same directory as the swf file automatically?

Hi all,
I'm really new here with flash and actionscript 3, so i might not make any sense, probably. Any input is greatly appreciated.
So here's my problem, i'm trying to make a Top Score history for a flash game, which i load from an XML file. The problem comes to when i want to update the XML file. how do i do that automatically without prompting the user (ofcourse). I want it to be saved automatically to the same directory as the flash file.
here's my code right now, but it prompts the user where it wants to be saved, which i don't want to happen:
var xml:XML = <XML>
                    <topScoreList>
                    </topScoreList>
              </XML>;
for(var i:Number=0; i<10; i++)
     var Record:XML = <Record>
                         <score></score>
                         <playerName></playerName>
                    </Record>;
     Record.score.appendChild(topScores[i].toString()); //topScores[] -> class variable containing scores
     Record.playerName.appendChild(topNames[i].toString()); //topNames[] -> class variable containing names
     xml.topScoreList.appendChild(Record);
var ba:ByteArray = new ByteArray();
ba.writeUTFBytes(xml);
var fr:FileReference = new FileReference();
fr.save(ba, "topScoreList.xml");

yeah that makes sense, actually - after reading a dozen more forums. It's really weird tho that flash doesn't have a Buffer writer, unlike other programming frameworks.
But yeah, thanks. i'll start reading on flash.net.SharedObject, correct ? Thanks.
Anyways, if anyone out there has a different input, will be greatly appreciated!!

How do i merge all the files in a directory into a single file

I need to merge(or move) all the files(all are .txt files) in a directory into a single file?

Thanks Martin, but I was searching for an inbuilt class or a stream like the SequenceInput class that could do this for me.

Get directory path. Just the directory path, without the file name.

Hi,
I need to export a file from a table in SAP to a directory in the PC. So i need to get a directory path.
I've been reading the threads but all the answers refer to functions F4_FILENAME, WS_FILENAME_GET, KD_GET_FILENAME_ON_F4 or to the method CL_GUI_FRONTEND_SERVICES=>FILE_SAVE_DIALOG.
But all of them require to select a file to close the window.
What I need is to get the path to the directory, not to a file.
For example:
I just need to get:
C:\TEMP
and not:
C:\TEMP\file.doc
Please help.
Thanks.

Pablo, use cl_gui_frontend_services=>directory_browse.
Refer this code:
PARAMETERS:
p_path         TYPE rlgrap-filename OBLIGATORY.
DATA:
gv_path_ini    TYPE string,
gv_path_sel    TYPE string.
AT SELECTION-SCREEN ON VALUE-REQUEST FOR p_path.
CALL METHOD cl_gui_frontend_services=>directory_browse
    EXPORTING
      initial_folder = gv_path_ini
    CHANGING
      selected_folder = gv_path_sel
    EXCEPTIONS
      cntl_error      = 1
      error_no_gui    = 2
      OTHERS          = 3
IF sy-subrc = 0.
    CALL METHOD cl_gui_cfw=>flush( ).
    IF NOT gv_path_sel IS INITIAL.
      gv_path_ini = p_path = gv_path_sel.
    ENDIF.
ENDIF.
As a side note, always avoid any WS_* function modules as they're not Unicode compliant.
Regards

Setting up the class root directory and choosing class files.

I made a simple test application as it is proposed at the J2EE 1.4 Tutorial and all worked.
(Chapter 24 Getting started with Enterprise Beans)
Than I deleted the ear file to try out the deploy mechanism again.
And after generating the new application with File-> New application which worked well I also
tried to use the EJB Wizard feature of the J2EE deployment tool. After I had selected the appropriate EJB classes to add to my jar, I clicked NEXT and I got the following message:
The class (converter.Converter) could not be loaded:
Please consult online help for assistance in setting up the class root directory and choosing class files.
The online helps isnt very usefull for this type of problem.
Do you have a hint what to do ?

Hi san-deepu,
I couldn't reproduce the error you are having when I followed Ch. 24 tutorial in packaging the ear. Is there anymore information in deploytool's logfile? This is in <user_home>/.deploytool/logfile, or you can run in verbose mode: <as_install>/bin/deploytool -v
When you say you deleted the ear file, did you also close the ear file in deploytool first? You may also want to exit deploytool () , and try deleting the temporary files. Deploytool usually cleans up the temp files automatically upon exit - maybe there are some left behind that it couldn't delete. On windows the temporary files are located by default in C:\Documents and Settings\Administrator\Local Settings\Temp\sun-dt-Administrator. In deploytool go to Edit --> Preferences --> General to find what the temporary directory is set to.
Which version of the appserver are you using? jdk version? operating system?
J

Directory Watcher is Skipping files

Similar Messages

Maybe you are looking for