Closed mixmix closed 1 year ago
Part | Duration |
---|---|
Create 5000 new messages | 357.73ms |
Validate 5000 messages | 476.85ms |
Native to db format 5000 messages | 109.40ms |
Db to native format 5000 messages | 85.78ms |
Add 1000 elements | 369.01ms |
Add 1000 box1 msgs | 1003.59ms |
Unbox 1000 box1 msgs first run | 217.60ms |
Unbox 1000 box1 msgs second run | 110.29ms |
Add 1000 box1 msgs | 962.72ms |
Query 1000 msgs first run | 48.74ms |
Query 1000 msgs second run | 12.27ms |
Add 1000 box2 msgs | 1284.96ms |
Unbox 1000 box2 msgs first run | 287.81ms |
Unbox 1000 box2 msgs second run | 134.50ms |
Migrate (+db1) | 10896.39ms |
Migrate (alone) | 3584.31ms |
Migrate (+db1 +db2) | 7916.77ms |
Migrate (+db2) | 5643.26ms |
Migrate continuation (+db2) | 820.19ms |
Memory usage without indexes | 755.58 MB = 40.69 MB + etc |
Initial indexing | 1539.53ms |
Initial indexing maxcpu=86 | 4716.76ms |
Initial indexing compat | 674.77ms |
Two indexes updating concurrently | 2728.11ms |
Key one initial | 46.44ms |
Key two | 5.35ms |
Key one again | 1.89ms |
Reboot and key one again | 47.23ms |
Latest root posts | 1699.01ms |
Latest posts | 7.80ms |
Votes one initial | 1435.17ms |
Votes again | 0.54ms |
HasRoot | 1283.60ms |
HasRoot again | 0.29ms |
Author one posts | 1385.78ms |
Author two posts | 15.59ms |
Dedicated author one posts | 1389.23ms |
Dedicated author one posts again | 0.53ms |
DeleteFeed | 5052.82ms |
Maximum memory usage | 987.53 MB = 58.88 MB + etc |
Indexes folder size | 10.01mb |
ok will keep tweaking. References:
oh I think I spotted my mistake, lets see
hmmm, error with Node 16 test
timing bug. re-running test
ITS BEAUTIFUL
Part | Duration |
---|---|
Create 5000 new messages | 372.37ms |
Validate 5000 messages | 459.54ms |
Native to db format 5000 messages | 97.53ms |
Db to native format 5000 messages | 86.84ms |
Add 1000 elements | 493.87ms |
Add 1000 box1 msgs | 978.91ms |
Unbox 1000 box1 msgs first run | 183.20ms |
Unbox 1000 box1 msgs second run | 94.85ms |
Add 1000 box1 msgs | 934.30ms |
Query 1000 msgs first run | 32.46ms |
Query 1000 msgs second run | 12.38ms |
Add 1000 box2 msgs | 1257.31ms |
Unbox 1000 box2 msgs first run | 279.93ms |
Unbox 1000 box2 msgs second run | 142.20ms |
Migrate (+db1) | 10246.56ms |
Migrate (alone) | 3129.76ms |
Migrate (+db1 +db2) | 7210.70ms |
Migrate (+db2) | 5080.99ms |
Migrate continuation (+db2) | 897.82ms |
Memory usage without indexes | 755.44 MB = 42.25 MB + etc |
Initial indexing | 1403.23ms |
Initial indexing maxcpu=86 | 4706.52ms |
Initial indexing compat | 593.20ms |
Two indexes updating concurrently | 2432.18ms |
Key one initial | 72.67ms |
Key two | 0.48ms |
Key one again | 0.91ms |
Reboot and key one again | 65.10ms |
Latest root posts | 1487.09ms |
Latest posts | 7.95ms |
Votes one initial | 1250.03ms |
Votes again | 0.60ms |
HasRoot | 1135.85ms |
HasRoot again | 0.39ms |
Author one posts | 1220.29ms |
Author two posts | 13.79ms |
Dedicated author one posts | 1210.38ms |
Dedicated author one posts again | 0.60ms |
DeleteFeed | 4939.17ms |
Maximum memory usage | 871.38 MB = 62.14 MB + etc |
Indexes folder size | 10.01mb |
the changes to sbot.close(t.end)
were because some tests failed with "but the server is already closed". This essentially drops all those errors (which I think it what we want errwhere?
(see image: https://github.com/ssbc/ssb-db2/pull/421#issuecomment-1504166429)
hmmm... passing, but didn't publish benchmark message?
Part | Duration |
---|---|
Create 5000 new messages | 351.05ms |
Validate 5000 messages | 436.91ms |
Native to db format 5000 messages | 102.89ms |
Db to native format 5000 messages | 89.39ms |
Add 1000 elements | 361.30ms |
Add 1000 box1 msgs | 899.60ms |
Unbox 1000 box1 msgs first run | 180.39ms |
Unbox 1000 box1 msgs second run | 88.99ms |
Add 1000 box1 msgs | 853.13ms |
Query 1000 msgs first run | 29.92ms |
Query 1000 msgs second run | 12.48ms |
Add 1000 box2 msgs | 1270.84ms |
Unbox 1000 box2 msgs first run | 349.76ms |
Unbox 1000 box2 msgs second run | 163.07ms |
Migrate (+db1) | 10883.37ms |
Migrate (alone) | 3587.81ms |
Migrate (+db1 +db2) | 7520.60ms |
Migrate (+db2) | 5544.26ms |
Migrate continuation (+db2) | 972.62ms |
Memory usage without indexes | 746.86 MB = 34.87 MB + etc |
Initial indexing | 1413.61ms |
Initial indexing maxcpu=86 | 4614.77ms |
Initial indexing compat | 710.23ms |
Two indexes updating concurrently | 2415.30ms |
Key one initial | 77.98ms |
Key two | 1.14ms |
Key one again | 1.21ms |
Reboot and key one again | 70.97ms |
Latest root posts | 1525.75ms |
Latest posts | 8.98ms |
Votes one initial | 1281.05ms |
Votes again | 0.34ms |
HasRoot | 1117.15ms |
HasRoot again | 0.31ms |
Author one posts | 1219.08ms |
Author two posts | 14.43ms |
Dedicated author one posts | 1211.33ms |
Dedicated author one posts again | 0.60ms |
DeleteFeed | 5250.22ms |
Maximum memory usage | 857.33 MB = 56.43 MB + etc |
Indexes folder size | 10.01mb |
I saw CI complaining:
So I took a BLIND STAB at updating the actions ... lets see if they run or if we hit breaking changes to API.