You can contact the technical support of the CESNET Storage Department via email:
support(at)cesnet.cz

Problem description I have forgotten my password which I was using to access my data storage at CESNET (password set up via VO application form).

Solution: Password can be set by using Perun system. By using this form you will change the password for most of virtual organizations in the e-infrastructure e.g. MetaCentrum password as well. To access this form, you have to use eduID.cz federation.

If you are applying for VO storage and registration process didn't require setting your password, you have probably an existing account in MetaCentrum. If you have forgotten your primary password, you can change it via link above.

Problem description Why does my connection timeout?

Solution: CESNET Data storage are using HSM (Hierarchical Storage Management) architecture. Within this HSM architecture can happen after some period of inactivity, that your data has been migrated to lower (slower) tiers. This leads to slower manipulation with your data, because before manipulation your data has to be migrated back to the higher (faster) tier. These delays can cause an timeout.

For successful download of your files, please set up the value of timeout for certain application. In some applications there may appear repeated request to enter your password. In this case please check the box with Remember my password.

Problem description How to consolidate more identities into one?

Solution: In case you have more than one identity from the various organizations and you wish to use for access to user management system (Perun) any one of them, you need to consolidate all those identities. From the side of user management system (Perun) it will look like one identity no matter which one you have used for login.

Method how to consolidate two fictitious identities - 4352@cvut.cz a jura03@fel.cvut.cz is described below. In case you have more than two identities you can repeat this method according to your needs.

1. Go to https://einfra.cesnet.cz/fed/ic/
2. Log in via you preferred identity
3. Choose in the section Add new way of signing in… eduID.cz identity (or any other one in case of need)
4. Log in via chosen identity

In case no problem has occurred, now you can use 4352@cvut.cz or jura03@fel.cvut.cz to access Perun system. From the system point of view it will look like one the same identity. In case of any difficulties please feel free to contact us at support(at)cesnet.cz.

Problem description How can I log out from all federated services?

Solution: Logout from the application doesn't mean logout in context of eduID.cz federation. For certain time period it is possible to log in the application without requiring authentication (without using login and password). Therefore, if there are other people which can reach your computer, we strongly recommend to close your web browser. By closing the web browser, you can achieve logout from all applications you have visited by using federation eduID.cz.

Problem description Are you getting lost in plenty of protocols and services? Aren't you sure which protocol or service is suitable for you?

Solution: Protocols and services overview with recommendations

Problem description Are you getting lost in plenty of protocols and services? Aren't you sure which address you should use for particular service?

Solution:

Problem description When I have connected to my data storage, I see some directories, which I didn't create.

Solution: In your Home directory in data storage are some subdirectories, which represent migration policies. It is very important to become familiar with this directory structure before you start to upload any data.

Problem description How do I get information about my quota and occupied space?

Solution: To check your quota and the amount of stored data, you can use the Accounting service.

Problem description After I have uploaded data into my data storage, my data doesn't keep original modification time. Is it possible to solve it?

Solution: Unfortunately FTP protocol doesn't allow set up time of files (in its current specification). RFC 3659 defines the MDTM command that allows you to get time but not to change it. However some FTP servers and clients allow to change the time using this command. We use FTP server which support MDTM command. Unfortunately clients FileZilla, WinSCP or Total Commander this command don't support. If you want to use MDTM, you can use the command-line client lftp with the ftp:use-mdtm-overloaded option.

Problem description Can I attach more volumes with various krb5 identities?

Solution: Start from Kerberos 1.10 it is possible to use the folder to keep your identity tickets instead of the file itself, so you can have more active identities in the same time and you can switch between them. Usually there is created the file in /tmp/ (e.g. /tmp/krb5cc_0_dX1u6IH8jO). It is possible to change the path into the selected folder.

First of all we will prepare a folder, where we want to save identity tickets. We will create this folder in /run/, because this folder will be cleaned after reboot.

# mkdir -p /run/user/0/krb5cc
# chmod og-rwx /run/user/0/krb5cc

We will set up into the variable KRB5CCNAME path to the given folder.

# export KRB5CCNAME=DIR:/run/user/0/krb5cc

Now our folder for saving identity tickets is prepared. We can start to initialize the identities.

# kinit user@DOMAIN1
# kinit user@EINFRA

To write out all available identity tickets we can use commands klist -A or klist -l.

# klist -A
Ticket cache: DIR::/run/user/0/krb5cc/tkta0Hysv
Default principal: user@DOMAIN1

Valid starting     Expires            Service principal
9.6.2015 13:54:44  9.6.2015 23:54:38  krbtgt/DOMAIN1@DOMAIN1
	renew until 24/06/2015 13:54 

Ticket cache: DIR::/run/user/0/krb5cc/tkt5PDllX
Default principal: user@EINFRA

Valid starting     Expires            Service principal
9.6.2015 13:54:22  9.6.2015 23:54:16  krbtgt/EINFRA@EINFRA
	renew until 16/06/2015 13:54 


# klist -l
Principal name                 Cache name
--------------                 ----------
user@DOMAIN1                   DIR::/run/user/0/krb5cc/tkta0Hysv
user@EINFRA                    DIR::/run/user/0/krb5cc/tkt5PDllX

We can see, which identity is active by using the command klist.

# klist
Ticket cache: DIR::/run/user/0/krb5cc/tkta0Hysv
Default principal: user@DOMAIN1

Valid starting     Expires            Service principal
9.6.2015 13:54:44  9.6.2015 23:54:38  krbtgt/DOMAIN1@DOMAIN1
	renew until 24/06/2015 13:54

To switch the identity we can use command kswitch.

# kswitch -p user@EINFRA

# klist
Ticket cache: DIR::/run/user/0/krb5cc/tkt5PDllX
Default principal: user@EINFRA

Valid starting     Expires            Service principal
9.6.2015 13:54:22  9.6.2015 23:54:16  krbtgt/EINFRA@EINFRA
	renew until 16/06/2015 13:54

In case we want to attach two NFSv4 volumes with different identities, we can run the command mount with active first identity, then we can switch the identity via kswitch and attach second volume with active second identity.

# mount -o sec=krb5i,proto=tcp,port=2049,intr storage-jihlava1-cerit.metacentrum.cz:/ /mnt/storage-cerit
# kswitch -p user@EINFRA
# mount -o rw,nfsvers=4,hard,intr,sec=krb5i nfs.du4.cesnet.cz:~/ /mnt/storage-du4

Problem description I tried to connect to the new Data Storage using NFS protocol, but it was not working.

Solution: In case of Ostrava Data Storage you have to insert the line into krb5.conf for [domain_realm] for Ostrava Data Storage, similarly for other new Data Storages.

.du4.cesnet.cz = EINFRA-SERVICES

Problem description: I tried to connect via NFS protocol, but I've got following error report.

mount.nfs: Network is unreachable

Solution: The issue is caused by failed startup rpc-gssd service, see the service status:

systemctl status rpc-gssd

Condition: start condition failed at Thu 2018-05-06 06:12:22 CET; 20s ago
           └─ ConditionPathExists=/etc/krb5.keytab was not met

To resolve this issue, you can create empty file in the /etc folder via following command:

touch /etc/krb5.keytab

Now you can restart rpc-gssd service and remount your storage.

Internet Explorer 6 is using vulnerable SSLv3. By the reason of keeping the security, our servers require TLSv1.0 and higher. Thus it is necessary to use web browser, which doesn't use SSL, but TLS.

Problem description: How to send notification in FileSender to multiple recipients about storing a file?

Solution: Separate the email addresses of recipients by comma.

Problem description How to send multiple files?

Solution: You can choose between two choices to achieve sending multiple files at once. One choice is to create an archive using WinZip, tar, etc. The second choice is to use a testing version of FileSender (filesender2.cesnet.cz). Because this is the testing version we can't assure the behavior of this instance. If you find a bug or strange behavior, so contact us at support(at)cesnet.cz.

Problem description What are the FileSender requirements on your browser?

Solution: Your browser must support HTML5 or Adobe Flash technology (Using Adobe Flash you can upload file sizes only up to 2GB).

Problem description What to do if you receive following message in FileSender:

Solution:

We are sorry but if you want to use the FileSender service, your account must be verified.
For more information how to verify your account, please visit https://eduid.cz.

If you have account in the academic identity federation eduID.cz and the FileSender shows you still same message as is mentioned above, please visit this page https://filesender.cesnet.cz/simplesaml/module.php/core/frontpage_auth.php

then choose:

• Test configured authentication sources

then click on filesender.cesnet.cz

then pick your organization (institution)

now please send screenshot of the output screen to: support(at)cesnet.cz This page will look like fig. below.

Problem description How long is the maximum file expiration period in FileSender?

Solution: Current version of FileSender allows to store data along one month period.

Description of the problem: Is it possible to upload files in a different way than via the web interface, i.e. directly from measuring or computing machines?

Solution: Yes, if the system has the ability to use the command line and python version 3, it is possible to use the offered script directly in the FileSender interface. You can find this script after logging in to FileSender under your user account (authentication via eduID) in the “My Profile” section and the “Python CLI Client” section, where you can download 2 files - the python script itself and the configuration containing the API key as well. which we upload to the path of our home directory ~/.filesender/filesender.py.ini . The script can be used without a configuration file, but at least the API-key and “User identification” (format abcdefghijklmnopqrstuvwxyz0123456789@einfra.cesnet.cz) must then be used as parameters on the command line. (this identifier can be found in the system Perun tab user - authentication - Login in EINFRAID-PERSISTENT)

Sample upload with downloaded and correctly placed configuration file:

python3 filesender.py -r abcdefghijklmnopqrstuvwxyz0123456789@einfra.cesnet.cz uploaded.file1 uploaded.file2

Sample upload using an API key:

python3 filesender.py -a 1234567890123456789012345678901234567890123456789012345678901234567890 -u abcdefghijklmnopqrstuvwxyz0123456789@einfra.cesnet.cz -r e-mail@recipient.com uploaded.file1 uploaded.file2

Problem description Does Globus support IPv6?

Solution: It does, however the support of IPv6 is currently experimental. Due to this, we haven't configured IPv6 in Globus on our servers yet.

Based on new Terms of Service that are in effect from January 2018, the system of handling data on storage facilities has been changed in order to prevent the facilities to get full. The data has been classified into two categories, permanent archives and backups.

The system is applied to storage facilities put into operation since 2018. It doesn't affect storage facilities du2, and du3 where all data is considered to be a permanent archive.

Problem description: What is archive and backups? What are handling differences in the storage facilities?

Solution: Data of archive type have permanent value. On the storage facility, it is limited by set quotas (limiting data size and number of files).

Backup spaces, on the other hand, serve to keep temporal data that doesn't have to be stored for more than months. Storing backups on the storage facility is limited by time: after one year, storage facility operators are entitled to delete the old data. The user is of course notified by e-mail in advance. This should suit majority of users who use the facility for backups. With reasonable strategy, there is no reason to keep older data for backups.

Storage facility operator typically provides backup space to the users. For permanent archival data, reasonable explanation and justification of resource usage is requested. Please note that misuse of archival spaces for backup data is considered a violation against the Terms of Service.

You can find detailed specification of both data types in section 3 of the Terms of Service.

Problem description: How do I know whether my storage space is archival or backup one?

Solution: It is based on folder structure, please see the description of the folder structure.

Problem description: I have a backup space and I need to keep some archival data.

Solution: Kindly contact your Virtual organisation manager who'll arrange adding a necessary storage space with the storage facility operator.

Problem description: I got notification, that I should move my data, but I cannot log in to Data Storage.

Solution: That issue can be caused by two reasons. First is expired membership in particular virtual organization (VO) or expiration of data moving period.

In case of suspicion for VO membership expiration, please check your inbox and make sure, that you prolonged your membership in particular VO. If you didn't do so, you can easily submit prolongation request (link can be found in the email notification about membership prolongation). Your membership (access to the data storage) will be prolonged/activated until 60 min after request being approved by VO admin.

Did you exceed data moving period? Time slot for moving your data is two weeks (14 days) since you received first email notification. After that period is your access to the Data Storage restricted from the reason of moving other groups of users. In case you did not finish moving of your data, please contact Data Storage Administrators and they will provide you additional access to the Data Storage.

Problem description: I tried to transfer my data using Globus/rsync protocol, but I got problem with low permissions, see the log below:

Error (transfer)
      Endpoint: ducesnet#globusonline (d8eb370a-6d04-11e5-ba46-22000b92c6ec)
      Server: globus.du4.cesnet.cz:2811
      File: /~/VO_storage-cache_tape/backups/tests/fLLgaxLv2R1/4Hcy5vd98A0
      Command: RETR ~/VO_storage-cache_tape/backups/tests/fLLgaxLv2R1/4Hcy5vd98A0
      Message: Fatal FTP response
      ---
      Details: 500-Command failed. : globus_l_gfs_file_open  failed.\r\n500-globus_xio: Unable to open file  
      /exports/home/username/VO_storage-cache_tape/VO_storage-cache_tape/backups/tests/fLLgaxLv2R1/4Hcy5vd98A0\r\n500- 
      globus_xio:  System error in open: Permission denied\r\n500-globus_xio: A system  call failed: Permission denied\r\n500 End.\r\n

Solution: That issue can be solved by checking and setting correct permissions on desired files/directories directly on the data storage. In case you use any system from Windows family, you can use this tutorial.

Please connect to the data storage using ssh protocol.

ssh username@ssh.du4.cesnet.cz

Now go to the directory, which you wish to check. Here is obvious which files are causing the issue. It is test1 a testfile1.

username@store:~>cd VO_storage-cache_tape
username@store:~/VO_storage-cache_tape>ls -alh
total 84K
drwx-----T    4 username storage  82 May  3 09:57 .
drwxr-x--- 1378 root     storage 56K May  3 13:30 ..
d---------    2 username storage  10 May  3 09:56 test1
drwxr-xr-x    2 username storage  10 May  3 09:56 test2
----------    1 username storage   0 May  3 09:56 testfile1
-rw-r--r--    1 username storage   0 May  3 09:56 testfile2

If you wish to look for problematic files/directories in the directory tree, you can use following command:

# find -L . -perm +u-rwx,g-rwx,o-rwx

Alternatively you can use find command and simultaneously correct missing permissions, see below:

# find -L . -perm +u-rwx,g-rwx,o-rwx -exec chmod u+rX '{}' \;