VIEWS: 22 PAGES: 6 POSTED ON: 3/23/2011
EDEN DATA WEBSITE HELPFILE SERIES Transferring files over the web There are three common ways to transfer data files using the internet: as attachments to emails; using a Web Browser (such as Internet Explorer, Netscape, or Firefox); or using “File Transfer Protocol” or FTP software. Email attachments can be used to send files to another email recipient, but cannot be used to download mail from another computer. Most email software packages have a menu item or toolbar button allowing a user to ‘Add attachment’ which usually opens a dialogue box in which files for attachment can be highlighted/selected. Once attached, the file is then sent as part of the email. Though several file attachments can be sent at once, to several recipients is so required, there is usually a limit on file size (3-10MB depending on the email system). Furthermore, dispatching large numbers of files to a range of recipients is very labour intensive for the sender. Browsers can be used to download files using links to data files posted on webpages. These links can ‘point’ to files on any computer that is linked to the web, provided that the computer on which the files are stored is set up to respond to file requests sent over the internet (i.e. as a ‘web server’). Browsers can therefore bring files to a local computer, but cannot send them elsewhere. Links can, however, be posted on a webpage, or sent to recipients by email, allowing them to download files from a local computer, but, once again, only if it is set up as a web server. This arrangement is comparatively simple for anyone who is web-literate, and with an appearance as complex or as utilitarian as the webmaster desires. If, however, the files change, then so must the links be revised. Accordingly, web pages of links, such as those provided on the EDEN data pages tend to be the favoured means of providing access to fixed sets of files by multiple users. FTP software can be used to transfer files both to and from another (‘remote’) machine provided that the remote computer is configured as an FTP ‘server’. Once the FTP server software is installed, data files can be downloaded from, and ‘uploaded’ to the other computer, with both processes being controlled from the local machine. There are thus three ways to send data files: email attachment; FTP transfer; or sending a link allowing the recipient to use a browser to download a file. There are also three ways to receive a file: being sent an email attachment; downloading data files from a remote web server using a web browser link that is in a webpage, document or email message; and FTP transfer. Emails and browsers thus require active human intervention, to either send or receive, and cannot be used by a single operator to do both. FTP can used by a single operator to both send and receive. FTP transfer is also usually substantially quicker than the other methods. For this reason FTP tends to be the method of choice for regular two way transfer of large numbers of files, or file collections which change from time to time. Most researchers have experience of email attachments and downloading files using browser links. For many, however, FTP software, especially sending (‘uploading’) files to other machines, is less familiar. This Help File is intended as a brief guide to using FTP to transfer files to and from a central archive site, using commonly available FTP software. Another source of help is http://www.ftpplanet.com/ftpresources/basics.htm, - and a Gooooogle search for ‘FTP tutorial’ will list many more. There are two categories of FTP software: the relatively user friendly ‘Client’ and the much more complex ‘Server’. Clients can send or receive from servers, but not other clients. As a result, server software must be installed on a ‘remote’ computer, whilst the local computer can host either client or server software. All archives must therefore have server software in order to distribute files, whilst only client software is needed to upload to and download from server hosted archives. A single server can handle transfer requests from many clients. There are many FTP software packages – most of which operate in much the same way. A very common Client is WS-FTP (http://www.ipswitch.com, used to be free, now shareware) or FTP commander (http://www.internet-soft.com/ftpcomm.htm, free). A widely used Server is Cerberus (http://www.cerberusftp.com/, free). Using WS-FTP Client software to download files. Starting WS Client (and most others) produces two widows side by side: the left hand pane lists files on the local computer, the right hand side is empty. To connect to a remote computer, click on the connect button, which will bring up a dialog box into which ths remote computer’s details need to be entered. These details include a connection Name (e.g. EDENdata), a web address or URL address (e.g. ergodd.zoo.ox.ac.uk), a username (e.g. eden) and a password (edendata). There is also an option to save the password, so that the connection details are stored on the local machine for later use. Once the details are entered or called up, and the connect button is clicked, the files or folders on the remote computer should be listed in the right- hand pane. Folders can be expanded with a left mouse click. Specific folders can be opened automatically (in WS-FTP) by entering the appropriate folder name in the Startup tab. To download a file, highlight it in the right-hand pane and click on the arrow button that points towards the left hand pane. To upload a file onto the remote computer, highlight a file in the left hand pane, and click on the arrow button that points towards the right hand window. Note that the ‘file permissions’ are set on the FTP server. These identify what operations each user can execute – (copy, overwrite, delete, rename etc) and which folders can be uploaded to or downloaded from. As a result, it will not, for example, always be possible to upload files into a remote folder as it may not have the appropriate permissions set. The permissions also dictate which remote folders a particular user can display which means that users can be assigned folders that other users cannot see or access. Some FTP sites will allow anonymous login so that a username is not needed to transfer files. Usually the username must be set to ‘anon’ or ‘anonymous’, and the password is an email address. Theses details can be entered in the connection dialogue box, and save to a named profile if required. Using a Browser to download files from a FTP server By putting FTP:// at the beginning of a Web address instead of the more usual http:// Browsers can be used to download files from an FTP server. The browser address should be of the form ftp://username:password@computeraddress, for example: ftp://eden:firstname.lastname@example.org which will display the /pubdata and /ftpupload folders held on the EDEN data website. The site is set up so that browsers can be used only to download data from the /pubdata folder. Clicking on the /pubdata link should produce: Some browsers may need an additional click on the reload/refresh icon (the arrows to the right of the red cross) to display the contents of a new folder. In Internet Explorer, right-clicking on a file (as opposed to a folder which is displayed in Bold type) and selecting ‘Save Target As’ will download the file. Note that only one file can be downloaded at a time, and http:// transfers are slower than FTP ones. If FTP doesn’t work, it is likely to be due to one of the following common problems: a) A firewall on the local or remote computer, or on one of the machines needed to operate an one of the host networks, blocks access in one or both directions. If the firewall on the local machine is not the culprit, little can be done by the ‘client’. Quite a lot of large institutions do no allow FTP archives or client installations on individual staff desktop machines on the grounds that security may not be as rigorously controlled as it could be. They do, however, usually provide centralised ftp facilities configured by network administrators. b) The wrong username or password is being used (these are usually case sensitive), or transfers are being attempted to/from locations/files which have not been assigned the proper permissions. Again the ‘client’ can do little to rectify this, bar trying to find the locations and permissions that match the username and passwords. c) The (remote) server is busy or constrained to service a specific number of clients at any one time. In addition, FTP servers are notoriously easy to crash and regularly need restarting. In which case a short wait and a second attempt may resolve a problem. The basics of setting up an FTP server Note that a server is only required on a machine hosting a data archive to which access by a number of users is required. Setting up an FTP server can seem very confusing, but in fact, once the software is installed, only consists of four major steps: setting up the data archive (not covered here); setting up a series of users with passwords; defining the locations of folders and files to which each user has access; and defining what operations each user can perform on the folders and files to which access has been provided. This overview will use Cerberus as an example. Setting up FTP users is similar to the process used to set up Windows users in that it must (in Windows 200 or later) be done by a user with Windows administrator privileges. Cerberus (and most other FTP servers) have a ‘User Manager’ utility accessible from a Configuration menu, which is used to define new users (below left, top left pane) and set whether access is via anonymous login, or needs a specific password (below left, top right pane). The bottom pane in the User Manager is used to identify which locations (files and folders) each user can access, and what operations can be performed within each folder. In the illustration below, the user eden does not have anonymous access, but needs a password; has been given access to folders /pubdata and /ftpupload and can only download files from /pubdata, but not upload them. Note that the folder /pubdata appears to have no other folders above it in the hierarchy. In actual fact, the full path of the folder /pubdata is c:\intepub\eden\pubdata. This rather confusing state of affairs is because FTP servers (and in fact Web servers as well) use ‘Virtual Root Directories’. The In this case the Cerberus server has assigned c:\intepub\eden\pubdata to be the top level (root) directory for user eden which therefore has access to all files and subfolders within /pubdata. Acees is not however permitted to files or folders above /pubdata (e.g. other files or folders in c:\inetpub or c:\). In this way folders from different levels in the real folder structure can be added to the virtual root directory (displayed here as \) to appear to be of the same level in the hierarchy. This is not only a way of restricting access to certain areas of the server, but also because the virtual directory structure can be separately configured for each user, access can be tailored and tightly controlled for each user. As far as the EDEN data website is concerned, this means that individual EDEN partners can be provided with very specific access rights for both download and upload – some general to all partners, some strictly limited to close collaborators – as requested.
Pages to are hidden for
"EDEN DATA WEBSITE HELPFILE SERIES Transferring files over the web "Please download to view full document