Re: [pbs-devel] [PATCH proxmox-backup 09/26] api2: admin: add unmount-device endpoint to datastore

[Hannes Laimer <h.laimer@proxmox.com>]

Am 06.07.22 um 13:43 schrieb Wolfgang Bumiller:
> On Tue, Jul 05, 2022 at 01:08:17PM +0000, Hannes Laimer wrote:
>> unmounts the removable device mounted at the datstore path and set the
>> maintenance-mode of the datastore to 'unplugged'. The worker waits for
>> all running tasks to finish before unmounting.
>>
>> Signed-off-by: Hannes Laimer <h.laimer@proxmox.com>
>> ---
>>   src/api2/admin/datastore.rs | 97 +++++++++++++++++++++++++++++++++----
>>   1 file changed, 87 insertions(+), 10 deletions(-)
>>
>> diff --git a/src/api2/admin/datastore.rs b/src/api2/admin/datastore.rs
>> index 5448fa10..2be95f92 100644
>> --- a/src/api2/admin/datastore.rs
>> +++ b/src/api2/admin/datastore.rs
>> @@ -22,27 +22,28 @@ use proxmox_router::{
>>       Router, RpcEnvironment, RpcEnvironmentType, SubdirMap,
>>   };
>>   use proxmox_schema::*;
>> +use proxmox_section_config::SectionConfigData;
>>   use proxmox_sys::fs::{
>>       file_read_firstline, file_read_optional_string, replace_file, CreateOptions,
>>   };
>> -use proxmox_sys::sortable;
>> -use proxmox_sys::{task_log, task_warn};
>> +use proxmox_sys::{sortable, task_log, task_warn, WorkerTaskContext};
>>   
>>   use pxar::accessor::aio::Accessor;
>>   use pxar::EntryKind;
>>   
>>   use pbs_api_types::{
>>       print_ns_and_snapshot, print_store_and_ns, Authid, BackupContent, BackupNamespace, BackupType,
>> -    Counts, CryptMode, DataStoreListItem, DataStoreStatus, GarbageCollectionStatus, GroupListItem,
>> -    KeepOptions, Operation, PruneJobOptions, RRDMode, RRDTimeFrame, SnapshotListItem,
>> -    SnapshotVerifyState, BACKUP_ARCHIVE_NAME_SCHEMA, BACKUP_ID_SCHEMA, BACKUP_NAMESPACE_SCHEMA,
>> -    BACKUP_TIME_SCHEMA, BACKUP_TYPE_SCHEMA, DATASTORE_SCHEMA, IGNORE_VERIFIED_BACKUPS_SCHEMA,
>> -    MAX_NAMESPACE_DEPTH, NS_MAX_DEPTH_SCHEMA, PRIV_DATASTORE_AUDIT, PRIV_DATASTORE_BACKUP,
>> -    PRIV_DATASTORE_MODIFY, PRIV_DATASTORE_PRUNE, PRIV_DATASTORE_READ, PRIV_DATASTORE_VERIFY,
>> -    UPID_SCHEMA, VERIFICATION_OUTDATED_AFTER_SCHEMA,
>> +    Counts, CryptMode, DataStoreConfig, DataStoreListItem, DataStoreStatus,
>> +    GarbageCollectionStatus, GroupListItem, KeepOptions, MaintenanceMode, MaintenanceType,
>> +    Operation, PruneJobOptions, RRDMode, RRDTimeFrame, SnapshotListItem, SnapshotVerifyState,
>> +    BACKUP_ARCHIVE_NAME_SCHEMA, BACKUP_ID_SCHEMA, BACKUP_NAMESPACE_SCHEMA, BACKUP_TIME_SCHEMA,
>> +    BACKUP_TYPE_SCHEMA, DATASTORE_SCHEMA, IGNORE_VERIFIED_BACKUPS_SCHEMA, MAX_NAMESPACE_DEPTH,
>> +    NS_MAX_DEPTH_SCHEMA, PRIV_DATASTORE_AUDIT, PRIV_DATASTORE_BACKUP, PRIV_DATASTORE_MODIFY,
>> +    PRIV_DATASTORE_PRUNE, PRIV_DATASTORE_READ, PRIV_DATASTORE_VERIFY, UPID_SCHEMA,
>> +    VERIFICATION_OUTDATED_AFTER_SCHEMA,
>>   };
>>   use pbs_client::pxar::{create_tar, create_zip};
>> -use pbs_config::CachedUserInfo;
>> +use pbs_config::{BackupLockGuard, CachedUserInfo};
>>   use pbs_datastore::backup_info::BackupInfo;
>>   use pbs_datastore::cached_chunk_reader::CachedChunkReader;
>>   use pbs_datastore::catalog::{ArchiveEntry, CatalogReader};
>> @@ -106,6 +107,39 @@ fn check_privs_and_load_store(
>>       Ok(datastore)
>>   }
>>   
>> +fn do_unmount_device(
>> +    _lock: BackupLockGuard,
>> +    mut config: SectionConfigData,
>> +    mut datastore: DataStoreConfig,
>> +    worker: Option<&dyn WorkerTaskContext>,
>> +) -> Result<(), Error> {
>> +    datastore.set_maintenance_mode(MaintenanceMode::new(MaintenanceType::Unplugged, None));
>> +    config.set_data(&datastore.name, "datastore", &datastore)?;
>> +    pbs_config::datastore::save_config(&config)?;
>> +    drop(_lock);
>> +
>> +    let mut active_operations = task_tracking::get_active_operations(&datastore.name)?;
>> +    while active_operations.read + active_operations.write > 0 {
> 
> I'd again prefer to fail than loop endlessly...
The idea was you start some sync/verify and then go to sleep or smth, so
it would unmount the device asap. But yes, we can also fail.
> 
>> +        if let Some(worker) = worker {
>> +            if worker.abort_requested() {
>> +                bail!("aborted, due to user request");
>> +            }
>> +            task_log!(
>> +                worker,
>> +                "can't unmount yet, still {} read and {} write operations active",
>> +                active_operations.read,
>> +                active_operations.write
>> +            );
>> +        }
>> +
>> +        std::thread::sleep(std::time::Duration::new(5, 0));
>> +        active_operations = task_tracking::get_active_operations(&datastore.name)?;
>> +    }
>> +
>> +    crate::tools::disks::unmount_by_mountpoint(&datastore.path)?;
>> +
>> +    Ok(())
>> +}
>>   fn read_backup_index(
>>       backup_dir: &BackupDir,
>>   ) -> Result<(BackupManifest, Vec<BackupContent>), Error> {