Re: [PATCH RFC 3/3] nfsd: Initial implementation of NFSv4 Courteous Server

[Chuck Lever III <chuck.lever@oracle.com>]

> On Feb 3, 2022, at 10:42 PM, Dai Ngo <dai.ngo@oracle.com> wrote:
> 
> 
> On 2/3/22 3:40 PM, Chuck Lever III wrote:
>> 
>>> On Feb 3, 2022, at 4:38 PM, Dai Ngo <dai.ngo@oracle.com> wrote:
>>> 
>>> 
>>> On 2/3/22 11:31 AM, Chuck Lever III wrote:
>>>>> On Jan 28, 2022, at 2:39 PM, Dai Ngo <dai.ngo@oracle.com> wrote:
>>>>> 
>>>>> Currently an NFSv4 client must maintain its lease by using the at least
>>>>> one of the state tokens or if nothing else, by issuing a RENEW (4.0), or
>>>>> a singleton SEQUENCE (4.1) at least once during each lease period. If the
>>>>> client fails to renew the lease, for any reason, the Linux server expunges
>>>>> the state tokens immediately upon detection of the "failure to renew the
>>>>> lease" condition and begins returning NFS4ERR_EXPIRED if the client should
>>>>> reconnect and attempt to use the (now) expired state.
>>>>> 
>>>>> The default lease period for the Linux server is 90 seconds.  The typical
>>>>> client cuts that in half and will issue a lease renewing operation every
>>>>> 45 seconds. The 90 second lease period is very short considering the
>>>>> potential for moderately long term network partitions.  A network partition
>>>>> refers to any loss of network connectivity between the NFS client and the
>>>>> NFS server, regardless of its root cause.  This includes NIC failures, NIC
>>>>> driver bugs, network misconfigurations & administrative errors, routers &
>>>>> switches crashing and/or having software updates applied, even down to
>>>>> cables being physically pulled.  In most cases, these network failures are
>>>>> transient, although the duration is unknown.
>>>>> 
>>>>> A server which does not immediately expunge the state on lease expiration
>>>>> is known as a Courteous Server.  A Courteous Server continues to recognize
>>>>> previously generated state tokens as valid until conflict arises between
>>>>> the expired state and the requests from another client, or the server
>>>>> reboots.
>>>>> 
>>>>> The initial implementation of the Courteous Server will do the following:
>>>>> 
>>>>> . When the laundromat thread detects an expired client and if that client
>>>>> still has established state on the Linux server and there is no waiters
>>>>> for the client's locks then deletes the client persistent record and marks
>>>>> the client as COURTESY_CLIENT and skips destroying the client and all of
>>>>> state, otherwise destroys the client as usual.
>>>>> 
>>>>> . Client persistent record is added to the client database when the
>>>>> courtesy client reconnects and transits to normal client.
>>>>> 
>>>>> . Lock/delegation/share reversation conflict with courtesy client is
>>>>> resolved by marking the courtesy client as DESTROY_COURTESY_CLIENT,
>>>>> effectively disable it, then allow the current request to proceed
>>>>> immediately.
>>>>> 
>>>>> . Courtesy client marked as DESTROY_COURTESY_CLIENT is not allowed to
>>>>> reconnect to reuse itsstate. It is expired by the laundromat asynchronously
>>>>> in the background.
>>>>> 
>>>>> Signed-off-by: Dai Ngo <dai.ngo@oracle.com>
>>>>> ---
>>>>> fs/nfsd/nfs4state.c | 454 +++++++++++++++++++++++++++++++++++++++++++++++-----
>>>>> fs/nfsd/state.h     |   5 +
>>>>> 2 files changed, 415 insertions(+), 44 deletions(-)
>>>>> 
>>>>> diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
>>>>> index 1956d377d1a6..b302d857e196 100644
>>>>> --- a/fs/nfsd/nfs4state.c
>>>>> +++ b/fs/nfsd/nfs4state.c
>>>>> @@ -1913,14 +1915,37 @@ __find_in_sessionid_hashtbl(struct nfs4_sessionid *sessionid, struct net *net)
>>>>> 
>>>>> static struct nfsd4_session *
>>>>> find_in_sessionid_hashtbl(struct nfs4_sessionid *sessionid, struct net *net,
>>>>> -		__be32 *ret)
>>>>> +		__be32 *ret, bool *courtesy_clnt)
>>>> IMO the new @courtesy_clnt parameter isn't necessary.
>>>> Just create a new cl_flag:
>>>> 
>>>> +#define NFSD4_COURTESY_CLIENT		(6)	/* be nice to expired client */
>>>> +#define NFSD4_DESTROY_COURTESY_CLIENT	(7)
>>>> 
>>>> #define NFSD4_CLIENT_PROMOTE_COURTESY   (8)
>>>> 
>>>> or REHYDRATE_COURTESY some such.
>>>> 
>>>> Set that flag and check it once it is safe to call
>>>> nfsd4_client_record_create().
>>> We need the 'courtesy_clnt' parameter so caller can specify
>>> whether the courtesy client should be promoted or not.
>> I understand what the flag is used for in the patch, but I
>> prefer to see this implemented without changing the synopsis
>> of all those functions. Especially adding an output parameter
>> like this is usually frowned upon.
>> 
>> The struct nfs_client can carry this flag, if not in cl_flags,
>> then perhaps in another field. That struct is visible in every
>> one of the callers.
> 
> The struct nfs4_client is not available to the caller of
> find_in_sessionid_hashtbl at the time it calls the function and
> the current input parameters of find_in_sessionid_hashtbl can
> not be used to specify this flag.

I see three callers of find_in_sessionid_hashtbl():

- nfsd4_bind_conn_to_session
- nfsd4_destroy_session
- nfsd4_sequence

In none of these callers is the courtesy_clnt variable set
to a true or false value _before_ find_in_sessionid_hashtbl()
is called. AFAICT, @courtesy_clnt is an output-only parameter.

The returned @session::se_client field points to a client
that can be examined to see if it has been promoted back to
active status.

--
Chuck Lever