| Message ID | 1655921718-28842-1-git-send-email-dai.ngo@oracle.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [1/1] NFSD: Fix memory shortage problem with Courteous server. | expand |
> On Jun 22, 2022, at 2:15 PM, Dai Ngo <dai.ngo@oracle.com> wrote: > > Currently the idle timeout for courtesy client is fixed at 1 day. > If there are lots of courtesy clients remain in the system it can > cause memory resource shortage that effects the operations of other > modules in the kernel. This problem can be observed by running pynfs > nfs4.0 CID5 test in a loop. Eventually system runs out of memory > and rpc.gssd fails to add new watch: > > rpc.gssd[3851]: ERROR: inotify_add_watch failed for nfsd4_cb/clnt6c2e: > No space left on device > > and also alloc_inode fails with out of memory: > > Call Trace: > <TASK> > dump_stack_lvl+0x33/0x42 > dump_header+0x4a/0x1ed > oom_kill_process+0x80/0x10d > out_of_memory+0x237/0x25f > __alloc_pages_slowpath.constprop.0+0x617/0x7b6 > __alloc_pages+0x132/0x1e3 > alloc_slab_page+0x15/0x33 > allocate_slab+0x78/0x1ab > ? alloc_inode+0x38/0x8d > ___slab_alloc+0x2af/0x373 > ? alloc_inode+0x38/0x8d > ? slab_pre_alloc_hook.constprop.0+0x9f/0x158 > ? alloc_inode+0x38/0x8d > __slab_alloc.constprop.0+0x1c/0x24 > kmem_cache_alloc_lru+0x8c/0x142 > alloc_inode+0x38/0x8d > iget_locked+0x60/0x126 > kernfs_get_inode+0x18/0x105 > kernfs_iop_lookup+0x6d/0xbc > __lookup_slow+0xb7/0xf9 > lookup_slow+0x3a/0x52 > walk_component+0x90/0x100 > ? inode_permission+0x87/0x128 > link_path_walk.part.0.constprop.0+0x266/0x2ea > ? path_init+0x101/0x2f2 > path_lookupat+0x4c/0xfa > filename_lookup+0x63/0xd7 > ? getname_flags+0x32/0x17a > ? kmem_cache_alloc+0x11f/0x144 > ? getname_flags+0x16c/0x17a > user_path_at_empty+0x37/0x4b > do_readlinkat+0x61/0x102 > __x64_sys_readlinkat+0x18/0x1b > do_syscall_64+0x57/0x72 > entry_SYSCALL_64_after_hwframe+0x46/0xb0 > RIP: 0033:0x7fce5410340e > > This patch adds a simple policy to dynamically adjust the idle > timeout based on the percentage of available memory in the system > as follow: > > . > 70% : unlimited. Courtesy clients are allowed to remain valid > as long as memory availability is above 70% > . 60% - 70%: 1 day. > . 50% - 60%: 1hr > . 40% - 50%: 30mins > . 30% - 40%: 15mins > . < 30%: disable. Expire all existing courtesy clients and donot > allow new courtesey client I thought our plan was to add a shrinker to do this. > Signed-off-by: Dai Ngo <dai.ngo@oracle.com> > --- > fs/nfsd/nfs4state.c | 41 +++++++++++++++++++++++++++++++++++++++-- > fs/nfsd/nfsd.h | 5 ++++- > 2 files changed, 43 insertions(+), 3 deletions(-) > > diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c > index 9409a0dc1b76..a7feea9d07cf 100644 > --- a/fs/nfsd/nfs4state.c > +++ b/fs/nfsd/nfs4state.c > @@ -5788,12 +5788,47 @@ nfs4_anylock_blockers(struct nfs4_client *clp) > return false; > } > > +static bool > +nfs4_allow_courtesy_client(struct nfsd_net *nn, unsigned int *idle_timeout) > +{ > + unsigned long avail; > + bool ret = true; > + unsigned int courtesy_expire = 0; > + struct sysinfo si; > + > + si_meminfo(&si); > + avail = (si.freeram * 10) / (si.totalram - si.totalhigh); > + switch (avail) { > + case 7: case 8: case 9: case 10: > + courtesy_expire = 0; /* unlimit */ > + break; > + case 6: > + courtesy_expire = NFSD_COURTESY_CLIENT_TO_1DAY; > + break; > + case 5: > + courtesy_expire = NFSD_COURTESY_CLIENT_TO_1HR; > + break; > + case 4: > + courtesy_expire = NFSD_COURTESY_CLIENT_TO_30MINS; > + break; > + case 3: > + courtesy_expire = NFSD_COURTESY_CLIENT_TO_15MINS; > + break; > + default: > + ret = false; /* disallow CC */ > + } > + *idle_timeout = courtesy_expire; > + return ret; > +} > + > static void > nfs4_get_client_reaplist(struct nfsd_net *nn, struct list_head *reaplist, > struct laundry_time *lt) > { > struct list_head *pos, *next; > struct nfs4_client *clp; > + unsigned int exptime; > + bool allow_cc = nfs4_allow_courtesy_client(nn, &exptime); > > INIT_LIST_HEAD(reaplist); > spin_lock(&nn->client_lock); > @@ -5803,11 +5838,13 @@ nfs4_get_client_reaplist(struct nfsd_net *nn, struct list_head *reaplist, > goto exp_client; > if (!state_expired(lt, clp->cl_time)) > break; > + if (!allow_cc) > + goto exp_client; > if (!atomic_read(&clp->cl_rpc_users)) > clp->cl_state = NFSD4_COURTESY; > if (!client_has_state(clp) || > - ktime_get_boottime_seconds() >= > - (clp->cl_time + NFSD_COURTESY_CLIENT_TIMEOUT)) > + (exptime && ktime_get_boottime_seconds() >= > + (clp->cl_time + exptime))) > goto exp_client; > if (nfs4_anylock_blockers(clp)) { > exp_client: > diff --git a/fs/nfsd/nfsd.h b/fs/nfsd/nfsd.h > index 847b482155ae..9d4a5708f852 100644 > --- a/fs/nfsd/nfsd.h > +++ b/fs/nfsd/nfsd.h > @@ -340,7 +340,10 @@ void nfsd_lockd_shutdown(void); > #define COMPOUND_ERR_SLACK_SPACE 16 /* OP_SETATTR */ > > #define NFSD_LAUNDROMAT_MINTIMEOUT 1 /* seconds */ > -#define NFSD_COURTESY_CLIENT_TIMEOUT (24 * 60 * 60) /* seconds */ > +#define NFSD_COURTESY_CLIENT_TO_1DAY (24 * 60 * 60) /* seconds */ > +#define NFSD_COURTESY_CLIENT_TO_1HR (60 * 60) > +#define NFSD_COURTESY_CLIENT_TO_30MINS (30 * 60) > +#define NFSD_COURTESY_CLIENT_TO_15MINS (15 * 60) > > /* > * The following attributes are currently not supported by the NFSv4 server: > -- > 2.9.5 > -- Chuck Lever
On 6/22/22 11:16 AM, Chuck Lever III wrote: > >> On Jun 22, 2022, at 2:15 PM, Dai Ngo <dai.ngo@oracle.com> wrote: >> >> Currently the idle timeout for courtesy client is fixed at 1 day. >> If there are lots of courtesy clients remain in the system it can >> cause memory resource shortage that effects the operations of other >> modules in the kernel. This problem can be observed by running pynfs >> nfs4.0 CID5 test in a loop. Eventually system runs out of memory >> and rpc.gssd fails to add new watch: >> >> rpc.gssd[3851]: ERROR: inotify_add_watch failed for nfsd4_cb/clnt6c2e: >> No space left on device >> >> and also alloc_inode fails with out of memory: >> >> Call Trace: >> <TASK> >> dump_stack_lvl+0x33/0x42 >> dump_header+0x4a/0x1ed >> oom_kill_process+0x80/0x10d >> out_of_memory+0x237/0x25f >> __alloc_pages_slowpath.constprop.0+0x617/0x7b6 >> __alloc_pages+0x132/0x1e3 >> alloc_slab_page+0x15/0x33 >> allocate_slab+0x78/0x1ab >> ? alloc_inode+0x38/0x8d >> ___slab_alloc+0x2af/0x373 >> ? alloc_inode+0x38/0x8d >> ? slab_pre_alloc_hook.constprop.0+0x9f/0x158 >> ? alloc_inode+0x38/0x8d >> __slab_alloc.constprop.0+0x1c/0x24 >> kmem_cache_alloc_lru+0x8c/0x142 >> alloc_inode+0x38/0x8d >> iget_locked+0x60/0x126 >> kernfs_get_inode+0x18/0x105 >> kernfs_iop_lookup+0x6d/0xbc >> __lookup_slow+0xb7/0xf9 >> lookup_slow+0x3a/0x52 >> walk_component+0x90/0x100 >> ? inode_permission+0x87/0x128 >> link_path_walk.part.0.constprop.0+0x266/0x2ea >> ? path_init+0x101/0x2f2 >> path_lookupat+0x4c/0xfa >> filename_lookup+0x63/0xd7 >> ? getname_flags+0x32/0x17a >> ? kmem_cache_alloc+0x11f/0x144 >> ? getname_flags+0x16c/0x17a >> user_path_at_empty+0x37/0x4b >> do_readlinkat+0x61/0x102 >> __x64_sys_readlinkat+0x18/0x1b >> do_syscall_64+0x57/0x72 >> entry_SYSCALL_64_after_hwframe+0x46/0xb0 >> RIP: 0033:0x7fce5410340e >> >> This patch adds a simple policy to dynamically adjust the idle >> timeout based on the percentage of available memory in the system >> as follow: >> >> . > 70% : unlimited. Courtesy clients are allowed to remain valid >> as long as memory availability is above 70% >> . 60% - 70%: 1 day. >> . 50% - 60%: 1hr >> . 40% - 50%: 30mins >> . 30% - 40%: 15mins >> . < 30%: disable. Expire all existing courtesy clients and donot >> allow new courtesey client > I thought our plan was to add a shrinker to do this. I'm not familiar with kernel's memory allocation and don't want to muck with it so I start with this simple approach but I'm open for any suggestion on how to add a shrinker for this task. Is there any existing model that I can use as reference? Thanks, -Dai > > >> Signed-off-by: Dai Ngo <dai.ngo@oracle.com> >> --- >> fs/nfsd/nfs4state.c | 41 +++++++++++++++++++++++++++++++++++++++-- >> fs/nfsd/nfsd.h | 5 ++++- >> 2 files changed, 43 insertions(+), 3 deletions(-) >> >> diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c >> index 9409a0dc1b76..a7feea9d07cf 100644 >> --- a/fs/nfsd/nfs4state.c >> +++ b/fs/nfsd/nfs4state.c >> @@ -5788,12 +5788,47 @@ nfs4_anylock_blockers(struct nfs4_client *clp) >> return false; >> } >> >> +static bool >> +nfs4_allow_courtesy_client(struct nfsd_net *nn, unsigned int *idle_timeout) >> +{ >> + unsigned long avail; >> + bool ret = true; >> + unsigned int courtesy_expire = 0; >> + struct sysinfo si; >> + >> + si_meminfo(&si); >> + avail = (si.freeram * 10) / (si.totalram - si.totalhigh); >> + switch (avail) { >> + case 7: case 8: case 9: case 10: >> + courtesy_expire = 0; /* unlimit */ >> + break; >> + case 6: >> + courtesy_expire = NFSD_COURTESY_CLIENT_TO_1DAY; >> + break; >> + case 5: >> + courtesy_expire = NFSD_COURTESY_CLIENT_TO_1HR; >> + break; >> + case 4: >> + courtesy_expire = NFSD_COURTESY_CLIENT_TO_30MINS; >> + break; >> + case 3: >> + courtesy_expire = NFSD_COURTESY_CLIENT_TO_15MINS; >> + break; >> + default: >> + ret = false; /* disallow CC */ >> + } >> + *idle_timeout = courtesy_expire; >> + return ret; >> +} >> + >> static void >> nfs4_get_client_reaplist(struct nfsd_net *nn, struct list_head *reaplist, >> struct laundry_time *lt) >> { >> struct list_head *pos, *next; >> struct nfs4_client *clp; >> + unsigned int exptime; >> + bool allow_cc = nfs4_allow_courtesy_client(nn, &exptime); >> >> INIT_LIST_HEAD(reaplist); >> spin_lock(&nn->client_lock); >> @@ -5803,11 +5838,13 @@ nfs4_get_client_reaplist(struct nfsd_net *nn, struct list_head *reaplist, >> goto exp_client; >> if (!state_expired(lt, clp->cl_time)) >> break; >> + if (!allow_cc) >> + goto exp_client; >> if (!atomic_read(&clp->cl_rpc_users)) >> clp->cl_state = NFSD4_COURTESY; >> if (!client_has_state(clp) || >> - ktime_get_boottime_seconds() >= >> - (clp->cl_time + NFSD_COURTESY_CLIENT_TIMEOUT)) >> + (exptime && ktime_get_boottime_seconds() >= >> + (clp->cl_time + exptime))) >> goto exp_client; >> if (nfs4_anylock_blockers(clp)) { >> exp_client: >> diff --git a/fs/nfsd/nfsd.h b/fs/nfsd/nfsd.h >> index 847b482155ae..9d4a5708f852 100644 >> --- a/fs/nfsd/nfsd.h >> +++ b/fs/nfsd/nfsd.h >> @@ -340,7 +340,10 @@ void nfsd_lockd_shutdown(void); >> #define COMPOUND_ERR_SLACK_SPACE 16 /* OP_SETATTR */ >> >> #define NFSD_LAUNDROMAT_MINTIMEOUT 1 /* seconds */ >> -#define NFSD_COURTESY_CLIENT_TIMEOUT (24 * 60 * 60) /* seconds */ >> +#define NFSD_COURTESY_CLIENT_TO_1DAY (24 * 60 * 60) /* seconds */ >> +#define NFSD_COURTESY_CLIENT_TO_1HR (60 * 60) >> +#define NFSD_COURTESY_CLIENT_TO_30MINS (30 * 60) >> +#define NFSD_COURTESY_CLIENT_TO_15MINS (15 * 60) >> >> /* >> * The following attributes are currently not supported by the NFSv4 server: >> -- >> 2.9.5 >> > -- > Chuck Lever > > >
> On Jun 22, 2022, at 2:28 PM, Dai Ngo <dai.ngo@oracle.com> wrote: > > > On 6/22/22 11:16 AM, Chuck Lever III wrote: >> >>> On Jun 22, 2022, at 2:15 PM, Dai Ngo <dai.ngo@oracle.com> wrote: >>> >>> Currently the idle timeout for courtesy client is fixed at 1 day. >>> If there are lots of courtesy clients remain in the system it can >>> cause memory resource shortage that effects the operations of other >>> modules in the kernel. This problem can be observed by running pynfs >>> nfs4.0 CID5 test in a loop. Eventually system runs out of memory >>> and rpc.gssd fails to add new watch: >>> >>> rpc.gssd[3851]: ERROR: inotify_add_watch failed for nfsd4_cb/clnt6c2e: >>> No space left on device >>> >>> and also alloc_inode fails with out of memory: >>> >>> Call Trace: >>> <TASK> >>> dump_stack_lvl+0x33/0x42 >>> dump_header+0x4a/0x1ed >>> oom_kill_process+0x80/0x10d >>> out_of_memory+0x237/0x25f >>> __alloc_pages_slowpath.constprop.0+0x617/0x7b6 >>> __alloc_pages+0x132/0x1e3 >>> alloc_slab_page+0x15/0x33 >>> allocate_slab+0x78/0x1ab >>> ? alloc_inode+0x38/0x8d >>> ___slab_alloc+0x2af/0x373 >>> ? alloc_inode+0x38/0x8d >>> ? slab_pre_alloc_hook.constprop.0+0x9f/0x158 >>> ? alloc_inode+0x38/0x8d >>> __slab_alloc.constprop.0+0x1c/0x24 >>> kmem_cache_alloc_lru+0x8c/0x142 >>> alloc_inode+0x38/0x8d >>> iget_locked+0x60/0x126 >>> kernfs_get_inode+0x18/0x105 >>> kernfs_iop_lookup+0x6d/0xbc >>> __lookup_slow+0xb7/0xf9 >>> lookup_slow+0x3a/0x52 >>> walk_component+0x90/0x100 >>> ? inode_permission+0x87/0x128 >>> link_path_walk.part.0.constprop.0+0x266/0x2ea >>> ? path_init+0x101/0x2f2 >>> path_lookupat+0x4c/0xfa >>> filename_lookup+0x63/0xd7 >>> ? getname_flags+0x32/0x17a >>> ? kmem_cache_alloc+0x11f/0x144 >>> ? getname_flags+0x16c/0x17a >>> user_path_at_empty+0x37/0x4b >>> do_readlinkat+0x61/0x102 >>> __x64_sys_readlinkat+0x18/0x1b >>> do_syscall_64+0x57/0x72 >>> entry_SYSCALL_64_after_hwframe+0x46/0xb0 >>> RIP: 0033:0x7fce5410340e >>> >>> This patch adds a simple policy to dynamically adjust the idle >>> timeout based on the percentage of available memory in the system >>> as follow: >>> >>> . > 70% : unlimited. Courtesy clients are allowed to remain valid >>> as long as memory availability is above 70% >>> . 60% - 70%: 1 day. >>> . 50% - 60%: 1hr >>> . 40% - 50%: 30mins >>> . 30% - 40%: 15mins >>> . < 30%: disable. Expire all existing courtesy clients and donot >>> allow new courtesey client >> I thought our plan was to add a shrinker to do this. > > I'm not familiar with kernel's memory allocation and don't want to muck > with it so I start with this simple approach but I'm open for any suggestion > on how to add a shrinker for this task. Is there any existing model that I > can use as reference? Fortunately there's nothing complicated about using a shrinker. Look for register_shrinker() calls to see code examples. There are two already in NFSD itself. -- Chuck Lever
On 6/22/22 11:32 AM, Chuck Lever III wrote: > >> On Jun 22, 2022, at 2:28 PM, Dai Ngo <dai.ngo@oracle.com> wrote: >> >> >> On 6/22/22 11:16 AM, Chuck Lever III wrote: >>>> On Jun 22, 2022, at 2:15 PM, Dai Ngo <dai.ngo@oracle.com> wrote: >>>> >>>> Currently the idle timeout for courtesy client is fixed at 1 day. >>>> If there are lots of courtesy clients remain in the system it can >>>> cause memory resource shortage that effects the operations of other >>>> modules in the kernel. This problem can be observed by running pynfs >>>> nfs4.0 CID5 test in a loop. Eventually system runs out of memory >>>> and rpc.gssd fails to add new watch: >>>> >>>> rpc.gssd[3851]: ERROR: inotify_add_watch failed for nfsd4_cb/clnt6c2e: >>>> No space left on device >>>> >>>> and also alloc_inode fails with out of memory: >>>> >>>> Call Trace: >>>> <TASK> >>>> dump_stack_lvl+0x33/0x42 >>>> dump_header+0x4a/0x1ed >>>> oom_kill_process+0x80/0x10d >>>> out_of_memory+0x237/0x25f >>>> __alloc_pages_slowpath.constprop.0+0x617/0x7b6 >>>> __alloc_pages+0x132/0x1e3 >>>> alloc_slab_page+0x15/0x33 >>>> allocate_slab+0x78/0x1ab >>>> ? alloc_inode+0x38/0x8d >>>> ___slab_alloc+0x2af/0x373 >>>> ? alloc_inode+0x38/0x8d >>>> ? slab_pre_alloc_hook.constprop.0+0x9f/0x158 >>>> ? alloc_inode+0x38/0x8d >>>> __slab_alloc.constprop.0+0x1c/0x24 >>>> kmem_cache_alloc_lru+0x8c/0x142 >>>> alloc_inode+0x38/0x8d >>>> iget_locked+0x60/0x126 >>>> kernfs_get_inode+0x18/0x105 >>>> kernfs_iop_lookup+0x6d/0xbc >>>> __lookup_slow+0xb7/0xf9 >>>> lookup_slow+0x3a/0x52 >>>> walk_component+0x90/0x100 >>>> ? inode_permission+0x87/0x128 >>>> link_path_walk.part.0.constprop.0+0x266/0x2ea >>>> ? path_init+0x101/0x2f2 >>>> path_lookupat+0x4c/0xfa >>>> filename_lookup+0x63/0xd7 >>>> ? getname_flags+0x32/0x17a >>>> ? kmem_cache_alloc+0x11f/0x144 >>>> ? getname_flags+0x16c/0x17a >>>> user_path_at_empty+0x37/0x4b >>>> do_readlinkat+0x61/0x102 >>>> __x64_sys_readlinkat+0x18/0x1b >>>> do_syscall_64+0x57/0x72 >>>> entry_SYSCALL_64_after_hwframe+0x46/0xb0 >>>> RIP: 0033:0x7fce5410340e >>>> >>>> This patch adds a simple policy to dynamically adjust the idle >>>> timeout based on the percentage of available memory in the system >>>> as follow: >>>> >>>> . > 70% : unlimited. Courtesy clients are allowed to remain valid >>>> as long as memory availability is above 70% >>>> . 60% - 70%: 1 day. >>>> . 50% - 60%: 1hr >>>> . 40% - 50%: 30mins >>>> . 30% - 40%: 15mins >>>> . < 30%: disable. Expire all existing courtesy clients and donot >>>> allow new courtesey client >>> I thought our plan was to add a shrinker to do this. >> I'm not familiar with kernel's memory allocation and don't want to muck >> with it so I start with this simple approach but I'm open for any suggestion >> on how to add a shrinker for this task. Is there any existing model that I >> can use as reference? > Fortunately there's nothing complicated about using a shrinker. > Look for register_shrinker() calls to see code examples. There > are two already in NFSD itself. Thanks Chuck, I'll take a look. -Dai > > > -- > Chuck Lever > > >
diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 9409a0dc1b76..a7feea9d07cf 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -5788,12 +5788,47 @@ nfs4_anylock_blockers(struct nfs4_client *clp) return false; } +static bool +nfs4_allow_courtesy_client(struct nfsd_net *nn, unsigned int *idle_timeout) +{ + unsigned long avail; + bool ret = true; + unsigned int courtesy_expire = 0; + struct sysinfo si; + + si_meminfo(&si); + avail = (si.freeram * 10) / (si.totalram - si.totalhigh); + switch (avail) { + case 7: case 8: case 9: case 10: + courtesy_expire = 0; /* unlimit */ + break; + case 6: + courtesy_expire = NFSD_COURTESY_CLIENT_TO_1DAY; + break; + case 5: + courtesy_expire = NFSD_COURTESY_CLIENT_TO_1HR; + break; + case 4: + courtesy_expire = NFSD_COURTESY_CLIENT_TO_30MINS; + break; + case 3: + courtesy_expire = NFSD_COURTESY_CLIENT_TO_15MINS; + break; + default: + ret = false; /* disallow CC */ + } + *idle_timeout = courtesy_expire; + return ret; +} + static void nfs4_get_client_reaplist(struct nfsd_net *nn, struct list_head *reaplist, struct laundry_time *lt) { struct list_head *pos, *next; struct nfs4_client *clp; + unsigned int exptime; + bool allow_cc = nfs4_allow_courtesy_client(nn, &exptime); INIT_LIST_HEAD(reaplist); spin_lock(&nn->client_lock); @@ -5803,11 +5838,13 @@ nfs4_get_client_reaplist(struct nfsd_net *nn, struct list_head *reaplist, goto exp_client; if (!state_expired(lt, clp->cl_time)) break; + if (!allow_cc) + goto exp_client; if (!atomic_read(&clp->cl_rpc_users)) clp->cl_state = NFSD4_COURTESY; if (!client_has_state(clp) || - ktime_get_boottime_seconds() >= - (clp->cl_time + NFSD_COURTESY_CLIENT_TIMEOUT)) + (exptime && ktime_get_boottime_seconds() >= + (clp->cl_time + exptime))) goto exp_client; if (nfs4_anylock_blockers(clp)) { exp_client: diff --git a/fs/nfsd/nfsd.h b/fs/nfsd/nfsd.h index 847b482155ae..9d4a5708f852 100644 --- a/fs/nfsd/nfsd.h +++ b/fs/nfsd/nfsd.h @@ -340,7 +340,10 @@ void nfsd_lockd_shutdown(void); #define COMPOUND_ERR_SLACK_SPACE 16 /* OP_SETATTR */ #define NFSD_LAUNDROMAT_MINTIMEOUT 1 /* seconds */ -#define NFSD_COURTESY_CLIENT_TIMEOUT (24 * 60 * 60) /* seconds */ +#define NFSD_COURTESY_CLIENT_TO_1DAY (24 * 60 * 60) /* seconds */ +#define NFSD_COURTESY_CLIENT_TO_1HR (60 * 60) +#define NFSD_COURTESY_CLIENT_TO_30MINS (30 * 60) +#define NFSD_COURTESY_CLIENT_TO_15MINS (15 * 60) /* * The following attributes are currently not supported by the NFSv4 server:
Currently the idle timeout for courtesy client is fixed at 1 day. If there are lots of courtesy clients remain in the system it can cause memory resource shortage that effects the operations of other modules in the kernel. This problem can be observed by running pynfs nfs4.0 CID5 test in a loop. Eventually system runs out of memory and rpc.gssd fails to add new watch: rpc.gssd[3851]: ERROR: inotify_add_watch failed for nfsd4_cb/clnt6c2e: No space left on device and also alloc_inode fails with out of memory: Call Trace: <TASK> dump_stack_lvl+0x33/0x42 dump_header+0x4a/0x1ed oom_kill_process+0x80/0x10d out_of_memory+0x237/0x25f __alloc_pages_slowpath.constprop.0+0x617/0x7b6 __alloc_pages+0x132/0x1e3 alloc_slab_page+0x15/0x33 allocate_slab+0x78/0x1ab ? alloc_inode+0x38/0x8d ___slab_alloc+0x2af/0x373 ? alloc_inode+0x38/0x8d ? slab_pre_alloc_hook.constprop.0+0x9f/0x158 ? alloc_inode+0x38/0x8d __slab_alloc.constprop.0+0x1c/0x24 kmem_cache_alloc_lru+0x8c/0x142 alloc_inode+0x38/0x8d iget_locked+0x60/0x126 kernfs_get_inode+0x18/0x105 kernfs_iop_lookup+0x6d/0xbc __lookup_slow+0xb7/0xf9 lookup_slow+0x3a/0x52 walk_component+0x90/0x100 ? inode_permission+0x87/0x128 link_path_walk.part.0.constprop.0+0x266/0x2ea ? path_init+0x101/0x2f2 path_lookupat+0x4c/0xfa filename_lookup+0x63/0xd7 ? getname_flags+0x32/0x17a ? kmem_cache_alloc+0x11f/0x144 ? getname_flags+0x16c/0x17a user_path_at_empty+0x37/0x4b do_readlinkat+0x61/0x102 __x64_sys_readlinkat+0x18/0x1b do_syscall_64+0x57/0x72 entry_SYSCALL_64_after_hwframe+0x46/0xb0 RIP: 0033:0x7fce5410340e This patch adds a simple policy to dynamically adjust the idle timeout based on the percentage of available memory in the system as follow: . > 70% : unlimited. Courtesy clients are allowed to remain valid as long as memory availability is above 70% . 60% - 70%: 1 day. . 50% - 60%: 1hr . 40% - 50%: 30mins . 30% - 40%: 15mins . < 30%: disable. Expire all existing courtesy clients and donot allow new courtesey client Signed-off-by: Dai Ngo <dai.ngo@oracle.com> --- fs/nfsd/nfs4state.c | 41 +++++++++++++++++++++++++++++++++++++++-- fs/nfsd/nfsd.h | 5 ++++- 2 files changed, 43 insertions(+), 3 deletions(-)