summaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* live: use .ant image as main image after image creation errorAlejandro Sirgo Rica2024-03-271-0/+6
| | | | | | | Restore image file from .ant to original file name if new image creation fails. Remove new imagen and move the .ant image file in place of the original as previously an error meant a rename of the image file without a revert to keep the image available.
* utils: make init_cache() use the cache mountpointAlejandro Sirgo Rica2024-03-261-1/+1
| | | | | | Make init_cache() use the actual cache mountpoint returned by the function mount_cache() for the creation of the cache directories instead of a hardcoded path.
* utils: remove redundant return statements from mount_mkdirAlejandro Sirgo Rica2024-03-261-3/+1
| | | | | | Remove return statement the program won't reach and simplify the return logic with a return from an if and a fallback return statement.
* utils: drop ogCopyEfiBootLoader scriptAlejandro Sirgo Rica2024-03-263-11/+53
| | | | | | | | Implement a Python equivalent of ogCopyEfiBootLoader as the function copy_efi_bootloader. This function copies the contents of the folder of the EFI loader in the ESP into a ogBoot folder at the root of the partition target of an image creation. copy_efi_bootloader is a Windows only functionality.
* live: improve lzop and partclone error handlingAlejandro Sirgo Rica2024-03-261-27/+47
| | | | | | | | | | | Control non 0 returncode of the lzop and partclone subprocess in image creation and restoration because this means that either lzop or partclone has failed. The implementation must cover cases such as not enough storage space and log errors into /tmp/command.log and the log file of the client handling the request. Check the returncode of lzop and partclone subprocesses and log the stderr of the process reporting non zero returncode.
* live: improve logging in image_createAlejandro Sirgo Rica2024-03-261-2/+2
| | | | | | Log the whole context of the error when an exception happens. The previous exception handling was hidding important information about the cause of the error.
* utils: add enforce_gpt argument to get_efi_partition()Alejandro Sirgo Rica2024-03-262-5/+8
| | | | | | | The Windows bootloader only supports a UEFI boot from a GPT partition. Set enforce_gpt to True in every codepath related to Windows. When enforce_gpt is set to True get_efi_partition() raises an exception when an MBR partition scheme is detected.
* revert 7f18485: improve uefi detection mechanismAlejandro Sirgo Rica2024-03-262-13/+3
| | | | | | | | | | | Make is_uefi_supported() only check for /sys/firmware/efi as get_efi_partition() will detect a missing ESP or an invalid partition scheme. Stop using get_efi_partition() inside is_uefi_supported() as the former is eventually called in every UEFI related code. UEFI supports both MBR and GPT as partition schemes and this is a required change to handle the particular case of Windows not being able to boot UEFI from a MBR partition scheme.
* src: make exception messages more contextual and explicitAlejandro Sirgo Rica2024-03-217-9/+9
| | | | | | | Provide more information in exception messages as those are the source of the logging messages. Add information about paths, files or configuration related to the operation associated to the exception.
* src: change generic exception types to be more explicitAlejandro Sirgo Rica2024-03-214-13/+13
| | | | | | | | Replace exception types to be more explicit about the nature of the error. Improve the exception raising semantics by using the 'from' keyword, this wraps an older exception into a new one so it is still considered the same object.
* utils: fix string not being defined as f-strings when they shouldAlejandro Sirgo Rica2024-03-211-2/+2
| | | | | Add the string prefix f'' to string instances formated as f-string but not properly constructed to expand the inline variables.
* virtual: add missing check=True to subprocess.run()Alejandro Sirgo Rica2024-03-211-2/+2
| | | | | Raise exception from subprocess.run() calls as previous code is trying to capture these exceptions to handle de errors.
* src: centralize error logging into send_internal_server_errorAlejandro Sirgo Rica2024-03-217-81/+53
| | | | | | | | | | | | | | | | | | | | Use only the exception messages as the main resource for error messages. The previous error code had string duplication in the form of: logging.error('msg here') raise Exception('msg here') That approach also has the downside of having log duplication as it had the local logging.err() and a global logging.exception() inside send_internal_server_error capturing the exception message. The actual code only requires raising an exception with a proper error message. Improve exception messages to give more error context. Log every AssertionError as a backtrace. Use the 'raise Exception from e' syntax to modify the a previously raised exception 'e' into an exception with aditional context or different type. This also prevents the message that warns about newer exceptions being launch after an initial exception.
* ogclient: consolidate logging in client coreAlejandro Sirgo Rica2024-03-211-3/+2
| | | | | | | | | The main function must be able to handle the login of critical error in the main ogClient class instance. Add a try except block to the ogCLient run logic and move the relevant error logs into the except block. Delegate the error messages to the exception message. This is the first step towards error message deduplication.
* utils: implement BIOS boot for windowsAlejandro Sirgo Rica2024-03-212-30/+10
| | | | | | | | | | | | | | | | | | Create ogboot.me and ogboot.secondboot as empty files and ogboot.firstboot with the value "iniciado" in the root of the BIOS Windows system partition. The files must contain data for GRUB to be able to write content, therefore these are created containing 3072 null bytes. The Windows boot process is handled by the "pxe" profile. There the files ogboot.me, ogboot.firstboot and ogboot.secondboot are used as a state machine to chose between booting Windows and ogLive. The first Windows boot happens if ogboot.me and ogboot.firstboot are identical, then "iniciado" is written in ogboot.firstboot. We skip this stage as we create ogboot.firstboot with 'iniciado'. The second Windows boot occurs if ogboot.me and ogboot.secondboot are boot identical, then "iniciado" is written in ogboot.secondboot. After the Windows boot ogLive is booted.
* utils: add disk index checks in get_partition_deviceAlejandro Sirgo Rica2024-03-211-1/+5
| | | | | Control the possibility of a bad disk index. Handle the checks in a similar fashion to the get_efi_partition function.
* utils: implement BIOS boot for LinuxAlejandro Sirgo Rica2024-03-212-1/+70
| | | | | | | | | Create a bios.py file to hold all the BIOS specific functions. Implement the _boot_bios_linux in Python. The new boot process tries to find the vmlinuz and initrd binaries at the desired partition. Then it tries to load them with kexec with the proper Grub boot params. One step closer to the removal of the boot legacy script.
* live: check if cache partition is available before calling tiptorrentAlejandro Sirgo Rica2024-03-211-0/+4
| | | | | | The image restore command must check if the cache partition is available. Otherwise if the user forgets to create the cache tiptorrent fails.
* utils: handle exceptions caused by the hivex packageAlejandro Sirgo Rica2024-03-081-3/+6
| | | | | | The mage creation process was being interrupted by an error trying to read the Windows registry by the Hivex library. Now the exceptions are handled and an error is reported.
* utils: implement linux distro id detectionAlejandro Sirgo Rica2024-03-082-1/+27
| | | | | | | The OS probe logic must be able to check a distro programmatically, add get_linux_distro_id to return an id whitout versioning. Ensure the availability of 'ubuntu' when we need to ensure certain features are only used with a supported system.
* utils: move all boot from OS functionality into boot.pyAlejandro Sirgo Rica2024-03-082-25/+51
| | | | | | | | This change is a preparative for reimplementing the BIOS boot in order to deprecate the legacy script. All the codepaths to boot systems located at a partition are now called from the boot_os_at function enabling an easier structure for the incoming code.
* utils: improve uefi detection mechanismAlejandro Sirgo Rica2024-03-083-4/+14
| | | | | | | Checking the existence /sys/firmware/efi as it might appear sometimes in BIOS installs if the BIOS configuration is not proper. Checking for the EFI partition is the safest method to veryfy the install type.
* utils: handle missing file in getlinuxversion functionAlejandro Sirgo Rica2024-03-081-11/+15
| | | | | | The function getlinuxversion receives a path to the os-release file. The case of not being able to open it was not handled and thus causing an unwanted exception.
* utils: improve logging in the get_efi_partition functionAlejandro Sirgo Rica2024-03-081-13/+16
| | | | | Log each partition that gets checked and make the exception messages more informative.
* uefi: update EFIBOOTMGR_BIN to use /opt/opengnsys/sbin/efibootmgrv1.3.2-6OpenGnSys Support Team2024-03-051-1/+1
| | | | debian package with json support provides the binary through this path, update it.
* uefi: define EFIBOOTMGR_BIN to ease changing the efibootmgr binaryAlejandro Sirgo Rica2024-03-041-9/+11
| | | | | | The json functionality proposed upstream might be merged one day in efibootmgr so deploying a fork would not be needed anymore. This change aims to ease the migration once that day comes.
* live: drop IniciarSesion script when uefi booting into linuxAlejandro Sirgo Rica2024-03-041-6/+33
| | | | | | Replace IniciarSesion script in favor of native Python code when booting a UEFI system into Linux. This completes the implementation of booting into an OS on a UEFI compliant system.
* live: drop IniciarSesion script when uefi bootingJose M. Guisado2024-03-041-0/+10
| | | | | | | | Replace IniciarSesion script in favor of native Python code when booting a UEFI system. This applies when running the "session" command. WIP: Only UEFI boots Windows systems. Raise NotImplementedError exception trying to boot a Linux system using UEFI.
* utils: add boot.pyJose M. Guisado2024-03-041-0/+63
| | | | | | | | | | Add utility module related to the process of booting a system from a client's partition. The main utility function to boot a clients system is boot_os_at(), from which firmware (UEFI or BIOS) and os-family specific private functions are invoked. This initial commit adds UEFI windows boot function.
* utils: add uefi.pyJose M. Guisado2024-03-041-0/+97
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add UEFI related utilities inside a new utility module: uefi.py _check_efibootmgr_json ====================== Check if the system efibootmgr executable supports JSON output. This is a private function used only by other functions from uefi.py. is_uefi_supported ================= Check if the system supports UEFI firmware. run_efibootmgr_json =================== Runs efibootmgr with json output support. Return the JSON output as a Python dict. efibootmgr_create_bootentry =========================== Create nvram boot entry. This bootentry is usually later set to boot next just once via "BootNext" nvram variable. efibootmgr_delete_bootentry =========================== Delete a nvram boot entry. Used to avoid duplicates when booting the same disk and partition from a given client. efibootmgr_bootnext =================== Set nvram "BootNext" variable to a given boot entry so after client reboot, PXE is not executed and the given boot entry takes precedence. Add dependency with efibootmgr version >= 18, and efibootmgr JSON output which is currently out of tree from util-linux repo.
* probe: add basic os family detectionJose M. Guisado2024-03-041-0/+16
| | | | | | | | Add a basic OS family enumeration: OSFamily. Add utility function that probes for an installed Linux or Windows system, returns the corresponding enum value, OSFamily.UNKNOWN otherwise.
* disk: add get_efi_partition functionJose M. Guisado2024-03-041-0/+24
| | | | | | | | | | | | | Add utility function inside disk.py to find, if any, the first ESP partition of a given disk. The disk is provided as an integer (starting at 1 following OpenGnsys scripts usual values), meaning the (n-1)th disk from the disk array returned from get_disks(). In the future a better mechanism should be put in place to fetch probed disks from a running client. This change is part of the upcoming drop of "IniciarSesion" script in favor of a Python native approach. Specifically regarding UEFI systems.
* cache: improve loggingOpenGnSys Support Team2024-02-221-1/+1
| | | | | use info instead of debug to make it easier to debug problems when creating the cache.
* live: improve logging with setup commandOpenGnSys Support Team2024-02-221-2/+10
| | | | | | Improve logging when setting up partition, provide more hints on progress. Fail in case partition layout is not supported.
* fs: add swap support and improve mkfs loggingOpenGnSys Support Team2024-02-221-3/+15
|
* utils: use returncode from subprocess in ogChangeRepo()OpenGnSys Support Team2024-02-211-4/+3
| | | | | | | ... the exception shows the samba password in the logs specify the error which tells us what has happened according to man mount(8) Return Codes.
* live: call partprobe on the specific diskOpenGnSys Support Team2024-02-191-1/+1
| | | | | otherwise partprobe does its best to find the disk, according to what I see through strace.
* poweroff: always call poweroff_oglive and _reboot_ogliveOpenGnSys Support Team2024-02-191-8/+2
| | | | Remove leftover fallback to directly call utilities to poweroff and reboot.
* fs: improve value parsing from command outputAlejandro Sirgo Rica2024-02-191-7/+29
| | | | | | value extraction did not have error checking and was handled in a one-liner. The actual implementation expands the parsing logic and moves it into a function.
* live: split logging to warn not to turn off client during image creationOpenGnSys Support Team2024-02-151-1/+2
| | | | just split this log message.
* utils: remove repository change in restoreImageCustomOpenGnSys Support Team2024-02-151-4/+0
| | | | this is broken, it uses default uses and password, remove it.
* fs: check if writing md5sum to full.sum file failsOpenGnSys Support Team2024-02-151-3/+11
| | | | writing to file might fail (permission denied, disk full), check for errors.
* fs: return unknown if blkid failsOpenGnSys Support Team2024-02-151-1/+7
| | | | instead of rising an exception
* fs: logging ext4 resize errorOpenGnSys Support Team2024-02-151-2/+4
| | | | log error in case resize2fs fails.
* fs: check if ntfsresize actually succeded to shrink filesystemOpenGnSys Support Team2024-02-151-1/+4
| | | | | According to ntfsresize.c, this retuns 0 in case nothing needs to be done. It should be safe to check for non-zero error and bail out in that case.
* fs: disentagle dry-run ntfsresize loop to probe for best shrink sizeOpenGnSys Support Team2024-02-151-8/+16
| | | | | | | | Revisit 5056b8f0d5ab ("fs: validate ntfsresize dry-run output") that has introduced a possible infinity loop. Disentangle this loop while at it: iterate until best smallest size is found by probing.
* src: ogChangeRepo returns zero on success and -1 on errorOpenGnSys Support Team2024-02-152-9/+12
| | | | | | | | | | | | | | | | | | | | do not return the returncode, instead return an integer. do not use except CalledProcessError as e: it causes a another exception while handling exception. Remount the original image repository. it should be possible to simplify this further by: - stacking mounts, no need to umount initial repo and mount it again when switching to the new repo, because remount back initial repo might fail (!) - use check=False and simply check for x.returncode
* live: remove mbuffer leftover in image restore commandOpenGnSys Support Team2024-02-151-1/+0
| | | | | | | | | | | | | | | | | | | Remove mbuffer, this is never used. mbuffer has been never been used since ogClient supports native image restore. Originally this was used like this: partclone [...] | mbuffer -q -M 40M | lzop [...] supposely to speed up partclone in case the device where the read happens is slowier than the device that is used for writes. See mbuffer(1) manpage examples. In any case, this needs benchmarking to really make sure this is helping. Remove it until that ever happens.
* live: log message improvements for image creation and restoreOpenGnSys Support Team2024-02-151-6/+12
| | | | | Provide more context information for debugging issues with image creation and restore.
* src: improve error check in image_create and image_restoreAlejandro Sirgo Rica2024-02-144-11/+29
| | | | | | | | | cover more error cases where exceptions need to be raised. check return code in the invoked subprocess. restoreImageCustom has been intentionally left behind, it is unclear what this custom script returns on success and error.