Jump to content
XPEnology Community

renegadeBE

Member
  • Posts

    232
  • Joined

  • Last visited

  • Days Won

    4

Posts posted by renegadeBE

  1. @Gastonzz

     

    Salut,

     

    Dans les dernières version d'arpl ou rr ce comportement est normal : le développeur à fait le choix de ne plus flasher l'adresse mac Custum directement sur le matériel. Il explique que pour faire fonctionner les différents services Synology, seul le S/n et la mac1 sont utilisés.

     

    Sont loader les renseignent différemment à présent à DSM. L(es) mac(s) physique(s) sont utilisées pour le reste. Ça évite certains problèmes liés au DHCP avec certains nics, mais aussi faciliter le fonctionnement de WoL.

     

    Pour t'assurer que celà fonctionne, utilise bien l'utilitaire des menus (je crois que c'est justement dans cmdline), tu devras alors rentrer le S/n et mac1.

     

    Bon amusement !

  2. @touyi

     

    Le reconstruction de ton chargeur n'arrangera pas ton soucis !

     

    Il faudrait reset tes paramètres réseaux ! J'ai dû bidouillé il y a un moment pour un problème de Firewall 😅.

     

    Il y a une option dans TCRP pour monter la partition système. Je regarderai ça ce week-end, pas assez de temps, il faut que je redécouvre moi aussi.

     

    Sinon, tu peux tenter une réinstallation de DSM. Tout sera mis par défaut, comptes utilisateurs également. Les données sont conservées.

     

    À mon avis @JacquesF saura te guider pour le re-parametrage via le terminal TCRP, il me semble calé dans tout ce qui est Linux 😇. J'espère qu'il fera un tour par ici.

  3. Salut à tous,

     

    Je ne sais pas, mais pourquoi s'obstiner à faire tourner 15 milles services sur une seule et même machine ? Attention, ne voyez aucune critique là, juste une réflexion !

     

    Mais, un nas à la base, est fait pour stocker et partager des fichiers, certe, dans un sens, et en extrapolant on peut voir le streaming comme un partage.

     

    J'ai longtemps réfléchi à changer mon architecture (Intel Atom) vers une machine plus puissante et donc, comme toi multiplier les services.

     

    Par la suite, je possédais une NVIDIA Shield 2019 de base (le cylindre), j'avais pour intension d'acheter une seconde pour ma salle de sport (le vélo est plus fun quand on a une petite série à se mettre sous la dent). En faisant quelques recherches, il existe un modèle Pro qui intègre un serveur PLEX. Après lecture sur le fonctionnement de ce dernier (j'étais accoutumé à Kodi), son déploiement, je me laissais de plus en plus tenté.

     

    Après analyses complémentaires, financièrement il était plus intéressant d'acheter ce modèle Pro. De plus il est tout à fait capable de transcoder plusieurs clients à la fois, leur petite puce est sacrément puissante et tournée vers le traitement d'images. Autre argument, sa consommation : quasi rien. En faisant quelques recherches tu retrouveras très facilement les détails.

     

    Plex accède aux fichiers (dossiers partagés) via Samba (compte créé expressément pour lui). Il partage ensuite la base de donnée sur le réseau. Il est accessible depuis l'extérieur par l'intermédiaire d'un reverse proxy, pas besoin de compte premium pour le transcodage, il est fourni gratuitement dans ce cas.

     

    Je te dépose ça là. Bonne réflexion 😌

  4. 14 minutes ago, Gronaldo said:

    This indicates a hardware issue, not related to the loader. Drives should allways spin up when machine starts/boots. 

     

    Hi,

     

    On startup, DSM loading, causes disk access (audible). In my case, no noise is emitted. I'm thinking more of a system crash (kernel panic?).

     

    I tried last night with the latest ARPL loader, direct boot mode.

    My server is fine and as usual, emitting disk access noises at startup.

     

    I was able to recover all the files and configurations.

     

    Do you think the problem will ever be solved? I have a preference for TCRP.

     

    THANKS :-)

  5. Hi @pocopico,

    I tried to migrate a test machine (amd 64bit platform): 7.1.1 42962U5 -> 7.2 64570.

    First, used shell commands: unable to create loader. The ALX module is incompatible with the requested version.

    Then I tried with the WEBUI. The process worked, the server was detected and I was able to install DSM 7.2. On reboot, the server was in recoverable mode. I applied, on reboot I launched the built-in FRIEND function to rebuild the loader. It has detected an update and has rebuilt the charger.

    After that, the server rebooted, but now the server is totally not found.

    If I rebuild the loader (without the minor update) it is in recovery mode again, after repair the server reboots. If I don't apply the "postupdate" procedure from the grub menu, I stay in the loop. If I apply the procedure from the GRUB menu, the server becomes invisible.

    The network card LEDs are both on, the server does not boot (no disk noise at boot). It displays the usual message on the screen...

     

    I specify that the loader was built in static mode.

     

    An idea ?

     

    EDIT : i use TCRP v0.9.4.9c

  6. @pocopico

     

    Feedback from a mac address issue that is not affected by the user_config.json settings:

     

    I had tried ARPL on a machine, I had customized the SN & MAC before the build of the loader. After that, the tests were not conclusive (AMD processor causing a Kernel Panic), having finished my tests, I returned to my loves: TCRP.

     

    When rebuilding the loader, I affected an S/N & MAC of a real Syno. At startup, in Synology Assistant I see that the old mac entered !!! Verification of GRUB: everything is OK, the mac address is the one entered when building the loader. Restart and again the wrong MAC...

     

    To solve the problem I had to flash a usb stick again with arpl, boot on it and customize the mac address.

     

    Upon reboot to TCRP, the new mac entered ARPL appeared. Also in Synology Assistant.

  7. On 6/18/2023 at 3:34 PM, Peter Suh said:

     

    There are 7 models that do not require PostUpdate.
    From the beginning, the full version of the DSM Pat file is compatible with Update 6.
    Models with a capacity of 300MB or more are applicable.
    https://archive.synology.com/download/Os/DSM/7.1.1-42962-6
    Redpill corresponds to the following 7 models.
    DS723+ DS923+ DS1823xs+ RS3621xs+ RS4021xs+ RS3618xs SA6400

     

    Although it is a Rackstation model among these models, the model most similar to DS3622xs+ is
    It is RS3621xs+. It has 12 bays and is the same Broadwell nk.
    We suggest temporarily migrating to this model until the issue with U6 is resolved in PostUpdate.
    M SHELL is currently working on a full version Update 6 and will be finished soon.

     

    2023-06-1811_19_39.thumb.png.039ee2d666760d2384443c7793826563.png

     

     

    2023-06-1811_19_14.thumb.png.d1137a2a837383658d08c5cd00f7625e.png

     

    Thx for your reply :-)

  8. Hi @pocopico

     

    I come report a issue with TCRP : The update 6 for ds3622xs 7.1.1-42962 break the loader (generated with the last TCRP). It causes an infinite recovery cycle... The command postupdate don't résolve this issue.

     

    When in doubt, I reinstalled the server, to recover it, I had to perform a clean installation (keep just data, no configs)! Otherwise, the loop continued.

     

    Once on version 7.1.1-42962-1, reconfigure all the parameters. I applied update 6 again, I switched to TCRP and launched the postupdate procedure, on restart, same problem. Return to TCRP to apply the procedure, but on restart, the server is still in the recovery cycle... :(

     

    Any ideas for solutions?

     

    THX :-)

     

    EDIT : My server uses an AMD platform, TCRP FRIEND does not work...

  9. @ginotator

     

    Tu étais dans une boucle de récupération.

     

    Après une mise à jour, le chargeur doit être reconstruit en partie. Pour réaliser cela, au reboot, il est impératif de faire un passage sur TinyCore et application la procédure "postupdate".

     

    Si  malencontreusement tu oubliais et te retrouvais à nouveau dans la situation, pas de panique !! Petit reboot et rdv sur TinyCore :-) (il n'est pas trop tard)

    Voici la liste des commandes avec explications :

     

    // Modèle émulé : DS3622xs (il faudra adapter la commande en fonction du modèle)
    // Mise à jour de l'utilitaire 
    
    ./rploader.sh fullupgrade
    ./rploader.sh update
    
    // On applique la procédure "postupdate"
    
    sudo ./rploader.sh postupdate ds3622xsp-7.1.1-42962 
    
    // Il va bosser et te demander de confirmer l'utilisation de la mise à jour
    
    Y -> ENTER
    
    // À nouveau, demande d'utilisation et d'appliquer
    
    Y -> ENTER
    
    // On redémarre
    
    exitcheck.sh reboot

     

    • Like 1
  10. Je vais essayer d'être le plus clair possible :-D,

     

    on commence avec la fameuse commande :

     

    sudo vi user_config.json

     

    L'éditeur se lance.

     

    Petit conseil : déplaces le curseur avec les flèches, place le au bon endroit, ensuite,

     

    on rentre en mode édition.

     

    Pour cela, plusieurs possibilités, pour ma part j'ai une préférence pour "i". Tu pousses donc sur le lettre "i" de ton clavier.

     

    Une fois que tu as édité le paramètre, on sort du mode édition avec ESC (escape).

     

    Tu déplaces à nouveau ton curseur sur la zone à éditer et on recommence : "i" -> édition -> "ESC"

     

    Une fois sorti du mode édition, on sauvegarde en tapant ":w" suivi "ENTER". On sort de vi ":x" suivi de "ENTER"

     

    Je résume :

     

    - sudo vi user_config.json

    - placement du curseur sur la zone à éditer

    - "i" (édition/insérer)

    - édition du paramètre

    - ESC (escape/sortir du mode édition)

    - déplacement du curseur avant l'édition d'un autre paramètre

    - "i" (édition/insérer)

    - édition du paramètre

    - ESC (escape/sortir du mode édition)

    - déplacement ....

    - etc.

    - ESC (escape/sortir du mode édition)

    - ":w" -> ENTER (Écrire le fichier)

    - ":x" -> ENTER (quitter en enregistrant)

     

     

    Donc, les commande vi pour l'édition

     

    i : insère des caractères après le curseur

    A : ajoute des caractères à la fin d’une ligne où que soit positionné le curseur

    o : insère une ligne après le curseur

    O : insère une ligne avant le curseur

    a : insère après le curseur

     

    Fichier

     

    :q! : quitter sans enregistrer

    :x : quitter en enregistrant

    :w : écrire/enregistrer le fichier

    :w nomdefichier : enregistrer sous

  11. @nono87

    Bon, c'est un peu overkill, mais tu peux simplement reconstruire ton chargeur (totalement), via TinyCore...

     

    Mais pour moi, il y a déjà un problème de mappage des disques, par défaut la valeur maxdisks est de 16... (et non 11 ou 12).
    Donc à mon avis, il y a un soucis dans sataportmap et ou diskidxmap.

     

    EDIT : (je suis un grand fan des modifications après relecture... hum)

     

    Je viens de lire ton dernier post, donc les 16 disques sont détectés. Bah au pire, reconstruis ton chargeur depuis TCRP... Tu pars de zéro, avant de lancer le build, prends soin de modifier via "vi" ton fichier user_config avec la valeur modifiée. Tu ne devras pas refaire l'installation de DSM, il devrait détecter les nouveau disques, tu devras juste les initialiser et les attacher à un volume ou en créer un nouveau.

     

    Il te faudra également modifier la valeur "internalportcfg": "0xffff" qui est la valeur hexadécimal de 16.

     

     "synoinfo": {
        "internalportcfg": "3fffff",
        "maxdisks": "22",
        "support_bde_internal_10g": "no",
        "support_disk_compatibility": "no",
        "support_memory_compatibility": "no"
      },

     

  12. - Outcome of the update: SUCCESSFUL

    - DSM version prior update: DSM 7.1.1-42962 Update 3 (Tinycore-Redpill v0.9.4.?)
    - Loader version and model: Tinycore-Redpill v0.9.4.3 - DS3622xs+
    - Using custom extra.lzma: NO
    - Installation type: BAREMETAL - AsRock Rack C2550D4I

    - Additional comments: Upgraded to DSM 7.1.1-42962 Update 5 via GUI & reboot to TinyCore

     

    Into TC, run the follow cmds :

     

    ./rploader.sh fullupgrade
    ./rploader.sh update 
    sudo ./rploader.sh postupdate ds3622xsp-7.1.1-42962 
    exitcheck.sh reboot

     

    • Like 1
    • Thanks 1
  13. Salut les gars,

     

    En regardant les détails de ta carte mère il s'agit bien du chiset H510 qui est utilisé.

     

    Dans tes configs, la remarque de nicoueron est bonne : vérifie si les contrôleurs sata sont bien en AHCI. De plus, toujours conseillé d'activité la fonction Hot Plug pour une bonne détection.

     

    L'installation de la plateforme de 920+ ne te donnera pas accès au transcodage matériel : ton processeur n'est pas pourvu d'un chip graphique intégré.

     

    J'ai l'impression qu'il y a un soucis de drivers (ou erreur de config bios).

     

    Si j'étais toi, j'essayerais avec le chargeur TCRP :

     

    Avantages :

    - évite l'installation d'une mise à jour non compatible par inadvertance,

    - plus de pilotes (supporte plus de matériel),

    - plus stable,

    - plus rapide,

    - outils complémentaire pour maintenance (montage partition système DSM),

    - moins de ressources consommées,

    - virtualisation via DSM fonctionnel.

     

    Inconvénients:

    - mappage des disques parfois laborieuse (à faire une seule fois),

    - il ne permet pas de mettre à jour automatiquement (installation manuel via GUI DSM, au redémarrage suivant : passage par l'utilitaire)

     

    Si tu as besoin d'une guidance dans les étapes, je peux t'y aider.

  14. Hi,

    After changing static ip address, I lost access to dsm, ssh, everything... My server is completely inaccessible. I forgot to temporarily disable the firewall (DSM). The firewall rules are very strict.

     

    Is there a way to reset network settings? Adding a comment in grub for it?

    Otherwise, I'll try forced_junior mode.

     

    Otherwise, mount the system partition (TinyCore) and modify the interface file, but I don't have the knowledge: it's 2 bonded ethernet in XOR bond. If I manage to modify the address in an accepted range, I will have access again.

     

    @Peter Suh@flyride@pocopico @IG-88?

     

    My bootloader is TCRP v9.4.x, dsm 7.1.1u3. No friend & Jun-mode option.

    Physical acces to the box : OK

     

    THANKS :)

  15. 5 minutes ago, pocopico said:

     

    We have notice that buildroot kexec causes kernel panic on some systems e.g. Microserver N54L which is also AMD based. Maybe

    you are also falling into the same issue. If thats true, then USB boot option is the only available solution. TCRP Friend , JUN mode and ARPL will not work as they are causing KP. 

     

    Okay, I will continue to update manually.

     

    Thank you for your answer :-)

     

     

  16. 9 hours ago, pocopico said:


    Well the IP is not a requirement for booting. It should go past that and boot. Do you mean you cannot get to the DSM ? What if you try to boot with the USB GRUB option ? 

     

    Indeed, DSM does not load.

     

    in USB mode works perfectly.

     

    When building the loader for the first time, it is important to declare the exact number of HDDs used: the motherboard has 6 SATA, only 3 are used -> satamap=3. Otherwise, the server would not start. Could the problem come from there? TCRP Friend use satamap parameters or autodetect ?

  17. 2 hours ago, pocopico said:

     

    First of all you can stop at TCRP Friend and check if network module is loaded and if the IP has been applied.

    lsmod
    lspci -k 
    ip a

     

     

     

     

    To try, I removed the arguments from the ip parameters. Values are default.

    It looks like it is getting the ip address from the router. Mac address entered in user_config.json is found in the eth0 interface.

     

    lsmod
    Module | Size | Used by | Not tainted
    alx     49152   0
    mdio    16384   1 alx

     

    lspci -k
    01:00.0 Network controller: Qualcomm Atheros AR9462 Wireless Network Adapter (rev 01)
            Subsystem: AzureWave AR9462 Wireless Network Adapter
    03:00.0 Ethernet controller: Qualcomm Atheros QCA8171 Gigabit Ethernet (rev 10)
            Subsystem ASRock Incorporation QCA8171 Gigabit Ethernet
            Kernel driver in use: alx
            Kernel modules: alx

     

    ip a
    1:  lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue qlen 1000
         link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
         inet 127.0.0.1/8 scope host lo
    2:  eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq qlen 1000
         link/ether XX:XX:XX:XX:XX:XX brd ff:ff:ff:ff:ff:ff
         inet 192.168.2.10/24 brd 192.168.2.255 scope global dynamic noprefixroute eth0
            valid_lfy ...........

     

    But, TCRP Friend does not start. Still the error message. (ERROR Could not get IP)

  18. Hi,

     

    I have a backup server. It is physically difficult to access. To make my life easier, I wanted to test TCRP Friend for this hardware.

    (MB: Asrock FM2A88X-ITX+, it uses qualcomm ethernet chip)

     

    It works fine with TCRP, but with the Friend option it never gets an ip! I get the error: ERROR Could not get IP. The server does not start.

     

    I tried with the default settings (DHCP), arbitrary configuration with my router(ip & mac binding), static option (edit user_config.json) and nothing is done! I tried using local dns server, cloudflare but nothing changes.

     

      "ipsettings": {
        "ipset": "static",
        "ipaddr": "192.168.2.10/24",
        "ipgw": "192.168.2.1",
        "ipdns": "1.1.1.1",
        "ipproxy": ""
      },

     

    I am using the latest version of rploader (v0.9.2.9). And TCRP Friend is in version 0.0.3! Platform : DS3622xs 7.1.1-42951 u2

     

    What can I do now?

     

    Thanks :D

     

    Sorry for my English :-)

×
×
  • Create New...