Skip to content

Chunked (large) item support #181

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 9 commits into from
Jul 13, 2016
Merged

Chunked (large) item support #181

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
0567967
chunked items checkpoint commit
dormando Jun 30, 2016
b05653f
chunked item second checkpoint
dormando Jul 1, 2016
51a828b
startup options for chunked items.
dormando Jul 7, 2016
3e10a71
protect binary sasl from chunked items.
dormando Jul 9, 2016
5978cf7
chunked append/prepend now work.
dormando Jul 9, 2016
6975235
CLRF in binprot was inverted.
dormando Jul 9, 2016
ee461d1
slabs reassigns works with chunks and chunked items.
dormando Jul 12, 2016
9c92c3a
update item_size_max tests
dormando Jul 12, 2016
5627405
fix test RAM bloat and more 32bit offsets.
dormando Jul 13, 2016
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
79 changes: 56 additions & 23 deletions items.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -61,6 +61,7 @@ static item *tails[LARGEST_ID];
static crawler crawlers[LARGEST_ID];
static itemstats_t itemstats[LARGEST_ID];
static unsigned int sizes[LARGEST_ID];
static uint64_t sizes_bytes[LARGEST_ID];
static unsigned int *stats_sizes_hist = NULL;
static uint64_t stats_sizes_cas_min = 0;
static int stats_sizes_buckets = 0;
Expand Down Expand Up @@ -89,7 +90,7 @@ void item_stats_reset(void) {
}

static int lru_pull_tail(const int orig_id, const int cur_lru,
const unsigned int total_chunks, const bool do_evict, const uint32_t cur_hv);
const uint64_t total_bytes, const bool do_evict, const uint32_t cur_hv);
static int lru_crawler_start(uint32_t id, uint32_t remaining);

/* Get the next CAS id for a new item. */
Expand Down Expand Up @@ -120,7 +121,7 @@ static unsigned int noexp_lru_size(int slabs_clsid) {
id |= NOEXP_LRU;
unsigned int ret;
pthread_mutex_lock(&lru_locks[id]);
ret = sizes[id];
ret = sizes_bytes[id];
pthread_mutex_unlock(&lru_locks[id]);
return ret;
}
Expand Down Expand Up @@ -162,7 +163,6 @@ item *do_item_alloc(char *key, const size_t nkey, const unsigned int flags,
uint8_t nsuffix;
item *it = NULL;
char suffix[40];
unsigned int total_chunks;
size_t ntotal = item_make_header(nkey + 1, flags, nbytes, suffix, &nsuffix);
if (settings.use_cas) {
ntotal += sizeof(uint64_t);
Expand All @@ -178,21 +178,34 @@ item *do_item_alloc(char *key, const size_t nkey, const unsigned int flags,
* occasional OOM's, rather than internally work around them.
* This also gives one fewer code path for slab alloc/free
*/
for (i = 0; i < 5; i++) {
/* TODO: if power_largest, try a lot more times? or a number of times
* based on how many chunks the new object should take up?
* or based on the size of an object lru_pull_tail() says it evicted?
* This is a classical GC problem if "large items" are of too varying of
* sizes. This is actually okay here since the larger the data, the more
* bandwidth it takes, the more time we can loop in comparison to serving
* and replacing small items.
*/
for (i = 0; i < 10; i++) {
uint64_t total_bytes;
/* Try to reclaim memory first */
if (!settings.lru_maintainer_thread) {
lru_pull_tail(id, COLD_LRU, 0, false, cur_hv);
}
it = slabs_alloc(ntotal, id, &total_chunks, 0);
it = slabs_alloc(ntotal, id, &total_bytes, 0);

if (settings.expirezero_does_not_evict)
total_chunks -= noexp_lru_size(id);
total_bytes -= noexp_lru_size(id);

if (it == NULL) {
if (settings.lru_maintainer_thread) {
lru_pull_tail(id, HOT_LRU, total_chunks, false, cur_hv);
lru_pull_tail(id, WARM_LRU, total_chunks, false, cur_hv);
lru_pull_tail(id, COLD_LRU, total_chunks, true, cur_hv);
lru_pull_tail(id, HOT_LRU, total_bytes, false, cur_hv);
lru_pull_tail(id, WARM_LRU, total_bytes, false, cur_hv);
if (lru_pull_tail(id, COLD_LRU, total_bytes, true, cur_hv) <= 0)
break;
} else {
lru_pull_tail(id, COLD_LRU, 0, true, cur_hv);
if (lru_pull_tail(id, COLD_LRU, 0, true, cur_hv) <= 0)
break;
}
} else {
break;
Expand All @@ -216,7 +229,8 @@ item *do_item_alloc(char *key, const size_t nkey, const unsigned int flags,
//assert(it != heads[id]);

/* Refcount is seeded to 1 by slabs_alloc() */
it->next = it->prev = it->h_next = 0;
it->next = it->prev = 0;

/* Items are initially loaded into the HOT_LRU. This is '0' but I want at
* least a note here. Compiler (hopefully?) optimizes this out.
*/
Expand All @@ -233,13 +247,29 @@ item *do_item_alloc(char *key, const size_t nkey, const unsigned int flags,
it->slabs_clsid = id;

DEBUG_REFCNT(it, '*');
it->it_flags = settings.use_cas ? ITEM_CAS : 0;
it->it_flags |= settings.use_cas ? ITEM_CAS : 0;
it->nkey = nkey;
it->nbytes = nbytes;
memcpy(ITEM_key(it), key, nkey);
it->exptime = exptime;
memcpy(ITEM_suffix(it), suffix, (size_t)nsuffix);
it->nsuffix = nsuffix;

/* Need to shuffle the pointer stored in h_next into it->data. */
if (it->it_flags & ITEM_CHUNKED) {
item_chunk *chunk = (item_chunk *) ITEM_data(it);

chunk->next = (item_chunk *) it->h_next;
chunk->prev = 0;
chunk->head = it;
/* Need to chain back into the head's chunk */
chunk->next->prev = chunk;
chunk->size = chunk->next->size - ((char *)chunk - (char *)it);
chunk->used = 0;
assert(chunk->size > 0);
}
it->h_next = 0;

return it;
}

Expand Down Expand Up @@ -288,6 +318,7 @@ static void do_item_link_q(item *it) { /* item is the new head */
*head = it;
if (*tail == 0) *tail = it;
sizes[it->slabs_clsid]++;
sizes_bytes[it->slabs_clsid] += it->nbytes;
return;
}

Expand Down Expand Up @@ -316,6 +347,7 @@ static void do_item_unlink_q(item *it) {
if (it->next) it->next->prev = it->prev;
if (it->prev) it->prev->next = it->next;
sizes[it->slabs_clsid]--;
sizes_bytes[it->slabs_clsid] -= it->nbytes;
return;
}

Expand Down Expand Up @@ -809,7 +841,7 @@ item *do_item_touch(const char *key, size_t nkey, uint32_t exptime,
/* Returns number of items remove, expired, or evicted.
* Callable from worker threads or the LRU maintainer thread */
static int lru_pull_tail(const int orig_id, const int cur_lru,
const unsigned int total_chunks, const bool do_evict, const uint32_t cur_hv) {
const uint64_t total_bytes, const bool do_evict, const uint32_t cur_hv) {
item *it = NULL;
int id = orig_id;
int removed = 0;
Expand All @@ -821,7 +853,7 @@ static int lru_pull_tail(const int orig_id, const int cur_lru,
item *next_it;
void *hold_lock = NULL;
unsigned int move_to_lru = 0;
uint64_t limit;
uint64_t limit = 0;

id |= cur_lru;
pthread_mutex_lock(&lru_locks[id]);
Expand Down Expand Up @@ -881,10 +913,11 @@ static int lru_pull_tail(const int orig_id, const int cur_lru,
*/
switch (cur_lru) {
case HOT_LRU:
limit = total_chunks * settings.hot_lru_pct / 100;
limit = total_bytes * settings.hot_lru_pct / 100;
case WARM_LRU:
limit = total_chunks * settings.warm_lru_pct / 100;
if (sizes[id] > limit) {
if (limit == 0)
limit = total_bytes * settings.warm_lru_pct / 100;
if (sizes_bytes[id] > limit) {
itemstats[id].moves_to_cold++;
move_to_lru = COLD_LRU;
do_item_unlink_q(search);
Expand Down Expand Up @@ -964,14 +997,14 @@ static int lru_maintainer_juggle(const int slabs_clsid) {
int i;
int did_moves = 0;
bool mem_limit_reached = false;
unsigned int total_chunks = 0;
uint64_t total_bytes = 0;
unsigned int chunks_perslab = 0;
unsigned int chunks_free = 0;
/* TODO: if free_chunks below high watermark, increase aggressiveness */
chunks_free = slabs_available_chunks(slabs_clsid, &mem_limit_reached,
&total_chunks, &chunks_perslab);
&total_bytes, &chunks_perslab);
if (settings.expirezero_does_not_evict)
total_chunks -= noexp_lru_size(slabs_clsid);
total_bytes -= noexp_lru_size(slabs_clsid);

/* If slab automove is enabled on any level, and we have more than 2 pages
* worth of chunks free in this class, ask (gently) to reassign a page
Expand All @@ -984,11 +1017,11 @@ static int lru_maintainer_juggle(const int slabs_clsid) {
/* Juggle HOT/WARM up to N times */
for (i = 0; i < 1000; i++) {
int do_more = 0;
if (lru_pull_tail(slabs_clsid, HOT_LRU, total_chunks, false, 0) ||
lru_pull_tail(slabs_clsid, WARM_LRU, total_chunks, false, 0)) {
if (lru_pull_tail(slabs_clsid, HOT_LRU, total_bytes, false, 0) ||
lru_pull_tail(slabs_clsid, WARM_LRU, total_bytes, false, 0)) {
do_more++;
}
do_more += lru_pull_tail(slabs_clsid, COLD_LRU, total_chunks, false, 0);
do_more += lru_pull_tail(slabs_clsid, COLD_LRU, total_bytes, false, 0);
if (do_more == 0)
break;
did_moves++;
Expand Down
Loading