Source:FAQ - Storage, Devices and Media Management
How do I use NAS/CIFS (QNAP, Synology ...) storage as backup target?
The below prerequisites must be met to use a NAS or a CIFS share as a backup storage. Depending on whether your SEP sesam Server or SEP sesam Remote Device Server (RDS) is Windows or Linux based, proceed as follows:
- To avoid double network traffic, connect the storage via an additional network card to the SEP sesam Server or RDS.
- If your SEP sesam Server or RDS is a Windows system, proceed as follows:
- Create a valid user in your AD/LDAP with administrative rights. For details, see Configuring LDAP/AD Authentication.
- Add this new user to your storage (NAS/CIFS) with full access to the storage part for SEP sesam.
- Adjust the SEP sesam system service and specify it to run under your newly created user account.
- Restart the service.
- Create a data store with a path, such as \\192.168.x.x\sepsesam.
- If your SEP sesam Server or RDS is a Linux system, create a permanent mount point to the server and use this mount point as a backup target for your data store.
How do I set up automatic archive adjustment?
You can set up automatic archive adjustment by creating a media event in the SEP sesam GUI.
By creating a special media event for archive adjustment, you can schedule it to run automatically at specific intervals or set up an archive adjustment template that allows you to start it interactively. Note that media events can be associated only with schedules that do not have any backup events configured.
- In the Main Selection -> Scheduling -> Schedules select a schedule without any backup events. Alternatively, click New Schedule and create a new schedule. For details, see Creating a Schedule.
- Right-click the selected schedule and then click New Media Event. Under Media action select Archive adjustment, and configure the parameters as required.
For more information and instructions, refer to Setting up Archive Adjustment.
Does SEP sesam support removable or USB media for backup?
There is limited support for removable disk media, including USB drives. For their configuration, see Configuring Removable Media.
How can I determine whether a tape is a SEP sesam tape or not?
You can check a tape label by using the SEP sesam GUI or a command line.
To check a tape label in the GUI, place the tape into a drive. Then select the drive in the GUI under Components -> Drives and execute the Identify label drive action. For details, see Checking and Labeling Tape Media. If the label is found, it will be displayed under the drive in the Current messages display.
Alternatively, go to the command line, switch to the directory <SESAM_ROOT>/bin/sms and execute the following depending on your operating system:
On Unix
./sm_sms_interface getlabel -d /dev/nst0
On Windows
sm_sms_interface getlabel -d Tape0
|
Tip |
| The necessary device (switch -d) can be found in GUI under Components -> Drives. |
How can I manually write a label on a tape (e.g. to overwrite an existing label without performing a check of the SEP sesam database)?
Go to the command line, switch to the <SESAM_ROOT>/bin/sms directory and execute the following depending on your operating system:
On Unix
./sm_sms_interface init -d /dev/nst0 -t testpool00001:1
On Windows
sm_sms_interface init -d Tape0 -t testpool00001:1
|
Warning |
| The tape will be overwritten without warning! |
The media label consists of three parts:
- Name of the pool (in the above example, the name is testpool).
- The tape ID, a 5-digit number.
- A number after the colon (':') that complies with the preceding number.
How do I automatically remove a tape after a backup is finished?
The best way of removing a tape after the backup is by using the sm_notify interface.
The standard unload command is:
sm_drive dismount 2 unload
This command refers to the selected drive, in this case drive 2.
Where do I install the cleaning tape?
When specifying Archive adjustment, the cleaning tape will automatically be loaded in sequence and an unwanted cleaning operation may be initiated. To prevent the cleaning operation from initiating, select the last (highest numbered) slot for the cleaning tape and reduce the number of slots by one (1).
Cleaning may take place by manually entering the cleaning tape or by following the manufacturer's instructions.
Why does SEP sesam lock tapes that were not involved in a backup?
This is related to full (FULL)->differential (DIFF) or FULL->incremental (INCR)->INCR->INCR-> ... backups.
The following example explains this problem:
- FULL -> requires three tapes (backup00001, backup00002, backup00003)
- INCR -> requires one tape (backup00004)
- INCR -> requires one tape (backup00005)
- INCR -> requires one tape (backup00006)
The first backup merely changes the EOL (end of lifetime) for the three tapes (backup00001 to backup00003), the second backup changes the EOL for the first four tapes (backup00001 to backup00004) and the third backup changes the EOL for the first five tapes (backup00001 to backup00005), and so on. For details on EOL, see Managing EOL.
If the next full backup fails, the succeeding incremental backup will use the chain from the last successful backup(s) and so the chain expands. This is necessary for a generation restore because an incremental backup includes only the changes from the last backup, regardless of whether it was a full, differential or incremental backup. All tapes are needed to perform a successful restore.
With a differential rather than an incremental backup, SEP sesam locks the tapes from full backup and the tapes that were used by the differential backup.
How do I get information on the oldest tape in a pool before the backup starts?
It is sometimes very useful to know which will be the next free medium in a media pool before a backup starts. To get this information, enter the following command:
sm_arch_getoldest 1 full-disk
The command shows the next free medium of pool full-disk in drive 1.
How do I check the consistency of a data store?
The following command allows you to check the consistency of a data store:
sm_data_store check_db -L {datastore} | -d {drive_num}
Example 1
The example shows a data store check that detects 2 savesets with the wrong data file size (sizes smaller than stored in the SEP sesam database). The state of both backup operations is either a fail ('X') or broken data transfer ('3') and the result status is therefore 'SUCCESS'.
#>sm_data_store check_db -L DATA_STORE2
2019-08-28 14:39:59: $Id: sm_data_store.py,v 1.105 2013/08/18 18:06:36 sesam Exp $
2019-08-28 14:39:59: Arguments: sm_data_store check_db -L DATA_STORE2
2019-08-28 14:39:59: $Id: sm_data_store.py,v 1.105 2013/08/18 18:06:36 sesam Exp $ Sesam Id: None
2019-08-28 14:39:59: Arguments: sm_data_store check_db -L DATA_STORE2
2019-08-28 14:39:59: DB: select * from data_stores where name= 'DATA_STORE2'
2019-08-28 14:39:59: DB: select * from hw_drives where data_store= 'DATA_STORE2' order by drive_num
2019-08-28 14:39:59: DB: select results.saveset,fdi_type,state,task,sesam_date,cnt,media_pool,results.uuid,data_size from results,result_lbls where
results.saveset= result_lbls.saveset and results.saveset in (select saveset from result_lbls where label in (select label from media where location= 'DATA_STORE2'))
and results.data_size>0 and results.state not in ('a','q') order by start_time
2019-08-28 14:39:59: DB: select results.saveset,fdi_type,state,task,sesam_date,cnt,media_pool,results.uuid,data_size from results,result_lbls where
results.saveset= result_lbls.saveset and results.saveset in (select saveset from result_lbls where label in (select label from media where location= 'DATA_STORE2'))
and results.data_size= 0 and results.state not in ('a','q') order by start_time
2019-08-28 14:39:59: DB: select label,pool from media where location= 'DATA_STORE2'
2019-08-28 14:39:59: Looking for savesets: sm_client dir "G:/DATA_STORE2/DATA_STORE2/DataStore1/DataStore100002"
2019-08-28 14:40:00: Looking for savesets: sm_client dir "G:/DATA_STORE2/DATA_STORE2/I_DS_POOL/I_DS_POOL00001"
2019-08-28 14:40:00: ==================================================================================
2019-08-28 14:40:00: Savesets in Sesam DB: 187.
2019-08-28 14:40:00: Saveset files found: 187.
2019-08-28 14:40:00: Saveset in DB and on FS: 187.
2019-08-28 14:40:00: Savesets not found on disk: 0.
2019-08-28 14:40:00: Saveset files not in DB: 0.
2019-08-28 14:40:00: -------------- List of savesets with wrong data file size ------------------------
2019-08-28 14:40:00: 3 SF20190826212205682@gp_o1hNIc8g miraculix_dev 2019-08-26 00:00:00 4430717952 > 4430686656
2019-08-28 14:40:00: X SC20190825083202946@iW9ZD7vcSw2 SESAM_BACKUP 2019-08-21 00:00:00 1323736064 > 1323385536
2019-08-28 14:40:00: -------------------- List of savesets without data file --------------------------
2019-08-28 14:40:00: -------------------- List of files without DB entry ------------------------------
2019-08-28 14:40:00: -------------------- List of files without data file -----------------------------
STATUS= SUCCESS MSG= OK
Example 2
A data store check that detects a saveset with the wrong data file size and with a successful backup state ('0'). The result status shows 'ERROR'.
G:\DATA_STORE2\DATA_STORE2\DataStore1\DataStore100002>sm_data_store check_db -L DATA_STORE2 2019-08-28 15:03:05: $Id: sm_data_store.py,v 1.105 2013/08/18 18:06:36 sesam Exp $ 2019-08-28 15:03:05: Arguments: sm_data_store check_db -L DATA_STORE2 ... 2019-08-28 15:03:06: Looking for savesets: sm_client dir "G:/DATA_STORE2/DATA_STORE2/DataStore1/DataStore100002" 2019-08-28 15:03:06: Looking for savesets: sm_client dir "G:/DATA_STORE2/DATA_STORE2/I_DS_POOL/I_DS_POOL00001" 2019-08-28 15:03:06: ================================================================================== 2019-08-28 15:03:06: Savesets in Sesam DB: 188. 2019-08-28 15:03:06: Saveset files found: 188. 2019-08-28 15:03:06: Saveset in DB and on FS: 188. 2019-08-28 15:03:06: Savesets not found on disk: 0. 2019-08-28 15:03:06: Saveset files not in DB: 0. 2019-08-28 15:03:06: -------------- List of savesets with wrong data file size ------------------------ 2019-08-28 15:03:06: 0! SI20190826150006895@ch-Gdm3Id_6 miraculix_dev 2019-08-26 00:00:00 106962944 > 1274932 2019-08-28 15:03:06: -------------------- List of savesets without data file -------------------------- 2019-08-28 15:03:06: -------------------- List of files without DB entry ------------------------------ 2019-08-28 15:03:06: -------------------- List of files without data file ----------------------------- 2019-08-28 15:03:06: ERROR: File size does not match: 1. STATUS= ERROR MSG= ERROR: File size does not match: 1.
In this case, the saveset is corrupted. You have to determine the reason for the corrupted saveset (disk IO problem, file system with limited file size, etc.). To avoid the corrupted saveset from being used, its result state must be changed to broken data transfer ('3').
sm_db "update results set state= '3', msg= 'corrupt saveset (size)' where saveset= 'SI20190826150006895@ch-Gdm3Id_6'"
If the data is still available on the backup source, a corresponding backup must be restarted.
Why is my single tape drive not in the list of devices when I try to initialize it?
First, inspect the device manually via OS options (for example Device Manager on Windows) to see if the tape drive is displayed. You can also check this by entering slu topology in the command line prompt, see Using slu topology for detecting devices.
If the device is not listed, the driver is not installed correctly. Look for the appropriate driver at the manufacturer's site or elsewhere on the internet, download it to your computer and run the program to install the driver. The media pool should now show the Tape (0). Finally, initialize the tapes. For details, see Media Strategy.
Why do I get the error message skipping blocks when I try to perform a SEP sesam backup with an Exabyte VXA172?
Some manufacturers often require a different driver for each tape model they manufacture. In this case, the VXA tape you initialized is not recognized by SEP sesam in the Exabyte tape drive. To fix this problem, go to the drive properties in the GUI (double-click the drive) and change the drive type from VXA to EXA.
Why does SEP sesam not write to my LTO drive on my x86 system?
The OS enables PAE (Physical Address Extension) automatically on x86 systems if more than 3.25 GB RAM are in use. With PAE, the OS cannot write more than 32 KB to a tape drive whereas SEP sesam needs 64 KB for LTO. The most efficient solution is to switch to an x64 OS. Decreasing the RAM to a maximum of 3.25 GB should also solve the problem.