2013-02-08 09:03:26 -08:00
|
|
|
// Copyright 2013 Prometheus Team
|
|
|
|
// Licensed under the Apache License, Version 2.0 (the "License");
|
|
|
|
// you may not use this file except in compliance with the License.
|
|
|
|
// You may obtain a copy of the License at
|
|
|
|
//
|
|
|
|
// http://www.apache.org/licenses/LICENSE-2.0
|
|
|
|
//
|
|
|
|
// Unless required by applicable law or agreed to in writing, software
|
|
|
|
// distributed under the License is distributed on an "AS IS" BASIS,
|
|
|
|
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
|
|
// See the License for the specific language governing permissions and
|
|
|
|
// limitations under the License.
|
|
|
|
|
|
|
|
package metric
|
|
|
|
|
|
|
|
import (
|
|
|
|
"testing"
|
|
|
|
"time"
|
2013-06-25 05:02:27 -07:00
|
|
|
|
|
|
|
clientmodel "github.com/prometheus/client_golang/model"
|
|
|
|
|
|
|
|
"github.com/prometheus/prometheus/utility/test"
|
2013-02-08 09:03:26 -08:00
|
|
|
)
|
|
|
|
|
|
|
|
func testBuilder(t test.Tester) {
|
|
|
|
type atTime struct {
|
|
|
|
fingerprint string
|
Use custom timestamp type for sample timestamps and related code.
So far we've been using Go's native time.Time for anything related to sample
timestamps. Since the range of time.Time is much bigger than what we need, this
has created two problems:
- there could be time.Time values which were out of the range/precision of the
time type that we persist to disk, therefore causing incorrectly ordered keys.
One bug caused by this was:
https://github.com/prometheus/prometheus/issues/367
It would be good to use a timestamp type that's more closely aligned with
what the underlying storage supports.
- sizeof(time.Time) is 192, while Prometheus should be ok with a single 64-bit
Unix timestamp (possibly even a 32-bit one). Since we store samples in large
numbers, this seriously affects memory usage. Furthermore, copying/working
with the data will be faster if it's smaller.
*MEMORY USAGE RESULTS*
Initial memory usage comparisons for a running Prometheus with 1 timeseries and
100,000 samples show roughly a 13% decrease in total (VIRT) memory usage. In my
tests, this advantage for some reason decreased a bit the more samples the
timeseries had (to 5-7% for millions of samples). This I can't fully explain,
but perhaps garbage collection issues were involved.
*WHEN TO USE THE NEW TIMESTAMP TYPE*
The new clientmodel.Timestamp type should be used whenever time
calculations are either directly or indirectly related to sample
timestamps.
For example:
- the timestamp of a sample itself
- all kinds of watermarks
- anything that may become or is compared to a sample timestamp (like the timestamp
passed into Target.Scrape()).
When to still use time.Time:
- for measuring durations/times not related to sample timestamps, like duration
telemetry exporting, timers that indicate how frequently to execute some
action, etc.
*NOTE ON OPERATOR OPTIMIZATION TESTS*
We don't use operator optimization code anymore, but it still lives in
the code as dead code. It still has tests, but I couldn't get all of them to
pass with the new timestamp format. I commented out the failing cases for now,
but we should probably remove the dead code soon. I just didn't want to do that
in the same change as this.
Change-Id: I821787414b0debe85c9fffaeb57abd453727af0f
2013-10-28 06:35:02 -07:00
|
|
|
time clientmodel.Timestamp
|
2013-02-08 09:03:26 -08:00
|
|
|
}
|
|
|
|
|
|
|
|
type atInterval struct {
|
|
|
|
fingerprint string
|
Use custom timestamp type for sample timestamps and related code.
So far we've been using Go's native time.Time for anything related to sample
timestamps. Since the range of time.Time is much bigger than what we need, this
has created two problems:
- there could be time.Time values which were out of the range/precision of the
time type that we persist to disk, therefore causing incorrectly ordered keys.
One bug caused by this was:
https://github.com/prometheus/prometheus/issues/367
It would be good to use a timestamp type that's more closely aligned with
what the underlying storage supports.
- sizeof(time.Time) is 192, while Prometheus should be ok with a single 64-bit
Unix timestamp (possibly even a 32-bit one). Since we store samples in large
numbers, this seriously affects memory usage. Furthermore, copying/working
with the data will be faster if it's smaller.
*MEMORY USAGE RESULTS*
Initial memory usage comparisons for a running Prometheus with 1 timeseries and
100,000 samples show roughly a 13% decrease in total (VIRT) memory usage. In my
tests, this advantage for some reason decreased a bit the more samples the
timeseries had (to 5-7% for millions of samples). This I can't fully explain,
but perhaps garbage collection issues were involved.
*WHEN TO USE THE NEW TIMESTAMP TYPE*
The new clientmodel.Timestamp type should be used whenever time
calculations are either directly or indirectly related to sample
timestamps.
For example:
- the timestamp of a sample itself
- all kinds of watermarks
- anything that may become or is compared to a sample timestamp (like the timestamp
passed into Target.Scrape()).
When to still use time.Time:
- for measuring durations/times not related to sample timestamps, like duration
telemetry exporting, timers that indicate how frequently to execute some
action, etc.
*NOTE ON OPERATOR OPTIMIZATION TESTS*
We don't use operator optimization code anymore, but it still lives in
the code as dead code. It still has tests, but I couldn't get all of them to
pass with the new timestamp format. I commented out the failing cases for now,
but we should probably remove the dead code soon. I just didn't want to do that
in the same change as this.
Change-Id: I821787414b0debe85c9fffaeb57abd453727af0f
2013-10-28 06:35:02 -07:00
|
|
|
from clientmodel.Timestamp
|
|
|
|
through clientmodel.Timestamp
|
2013-02-08 09:03:26 -08:00
|
|
|
interval time.Duration
|
|
|
|
}
|
|
|
|
|
|
|
|
type atRange struct {
|
|
|
|
fingerprint string
|
Use custom timestamp type for sample timestamps and related code.
So far we've been using Go's native time.Time for anything related to sample
timestamps. Since the range of time.Time is much bigger than what we need, this
has created two problems:
- there could be time.Time values which were out of the range/precision of the
time type that we persist to disk, therefore causing incorrectly ordered keys.
One bug caused by this was:
https://github.com/prometheus/prometheus/issues/367
It would be good to use a timestamp type that's more closely aligned with
what the underlying storage supports.
- sizeof(time.Time) is 192, while Prometheus should be ok with a single 64-bit
Unix timestamp (possibly even a 32-bit one). Since we store samples in large
numbers, this seriously affects memory usage. Furthermore, copying/working
with the data will be faster if it's smaller.
*MEMORY USAGE RESULTS*
Initial memory usage comparisons for a running Prometheus with 1 timeseries and
100,000 samples show roughly a 13% decrease in total (VIRT) memory usage. In my
tests, this advantage for some reason decreased a bit the more samples the
timeseries had (to 5-7% for millions of samples). This I can't fully explain,
but perhaps garbage collection issues were involved.
*WHEN TO USE THE NEW TIMESTAMP TYPE*
The new clientmodel.Timestamp type should be used whenever time
calculations are either directly or indirectly related to sample
timestamps.
For example:
- the timestamp of a sample itself
- all kinds of watermarks
- anything that may become or is compared to a sample timestamp (like the timestamp
passed into Target.Scrape()).
When to still use time.Time:
- for measuring durations/times not related to sample timestamps, like duration
telemetry exporting, timers that indicate how frequently to execute some
action, etc.
*NOTE ON OPERATOR OPTIMIZATION TESTS*
We don't use operator optimization code anymore, but it still lives in
the code as dead code. It still has tests, but I couldn't get all of them to
pass with the new timestamp format. I commented out the failing cases for now,
but we should probably remove the dead code soon. I just didn't want to do that
in the same change as this.
Change-Id: I821787414b0debe85c9fffaeb57abd453727af0f
2013-10-28 06:35:02 -07:00
|
|
|
from clientmodel.Timestamp
|
|
|
|
through clientmodel.Timestamp
|
2013-02-08 09:03:26 -08:00
|
|
|
}
|
|
|
|
|
|
|
|
type in struct {
|
|
|
|
atTimes []atTime
|
|
|
|
atIntervals []atInterval
|
|
|
|
atRanges []atRange
|
|
|
|
}
|
|
|
|
|
|
|
|
type out []struct {
|
|
|
|
fingerprint string
|
|
|
|
operations ops
|
|
|
|
}
|
|
|
|
|
|
|
|
var scenarios = []struct {
|
|
|
|
in in
|
|
|
|
out out
|
|
|
|
}{
|
2013-04-18 07:10:52 -07:00
|
|
|
// Ensure that the fingerprint is sorted in proper order.
|
2013-02-08 09:03:26 -08:00
|
|
|
{
|
|
|
|
in: in{
|
|
|
|
atTimes: []atTime{
|
|
|
|
{
|
|
|
|
fingerprint: "0000000000000001111-a-4-a",
|
Use custom timestamp type for sample timestamps and related code.
So far we've been using Go's native time.Time for anything related to sample
timestamps. Since the range of time.Time is much bigger than what we need, this
has created two problems:
- there could be time.Time values which were out of the range/precision of the
time type that we persist to disk, therefore causing incorrectly ordered keys.
One bug caused by this was:
https://github.com/prometheus/prometheus/issues/367
It would be good to use a timestamp type that's more closely aligned with
what the underlying storage supports.
- sizeof(time.Time) is 192, while Prometheus should be ok with a single 64-bit
Unix timestamp (possibly even a 32-bit one). Since we store samples in large
numbers, this seriously affects memory usage. Furthermore, copying/working
with the data will be faster if it's smaller.
*MEMORY USAGE RESULTS*
Initial memory usage comparisons for a running Prometheus with 1 timeseries and
100,000 samples show roughly a 13% decrease in total (VIRT) memory usage. In my
tests, this advantage for some reason decreased a bit the more samples the
timeseries had (to 5-7% for millions of samples). This I can't fully explain,
but perhaps garbage collection issues were involved.
*WHEN TO USE THE NEW TIMESTAMP TYPE*
The new clientmodel.Timestamp type should be used whenever time
calculations are either directly or indirectly related to sample
timestamps.
For example:
- the timestamp of a sample itself
- all kinds of watermarks
- anything that may become or is compared to a sample timestamp (like the timestamp
passed into Target.Scrape()).
When to still use time.Time:
- for measuring durations/times not related to sample timestamps, like duration
telemetry exporting, timers that indicate how frequently to execute some
action, etc.
*NOTE ON OPERATOR OPTIMIZATION TESTS*
We don't use operator optimization code anymore, but it still lives in
the code as dead code. It still has tests, but I couldn't get all of them to
pass with the new timestamp format. I commented out the failing cases for now,
but we should probably remove the dead code soon. I just didn't want to do that
in the same change as this.
Change-Id: I821787414b0debe85c9fffaeb57abd453727af0f
2013-10-28 06:35:02 -07:00
|
|
|
time: clientmodel.TimestampFromUnix(100),
|
2013-02-08 09:03:26 -08:00
|
|
|
},
|
|
|
|
{
|
|
|
|
fingerprint: "0000000000000000000-a-4-a",
|
Use custom timestamp type for sample timestamps and related code.
So far we've been using Go's native time.Time for anything related to sample
timestamps. Since the range of time.Time is much bigger than what we need, this
has created two problems:
- there could be time.Time values which were out of the range/precision of the
time type that we persist to disk, therefore causing incorrectly ordered keys.
One bug caused by this was:
https://github.com/prometheus/prometheus/issues/367
It would be good to use a timestamp type that's more closely aligned with
what the underlying storage supports.
- sizeof(time.Time) is 192, while Prometheus should be ok with a single 64-bit
Unix timestamp (possibly even a 32-bit one). Since we store samples in large
numbers, this seriously affects memory usage. Furthermore, copying/working
with the data will be faster if it's smaller.
*MEMORY USAGE RESULTS*
Initial memory usage comparisons for a running Prometheus with 1 timeseries and
100,000 samples show roughly a 13% decrease in total (VIRT) memory usage. In my
tests, this advantage for some reason decreased a bit the more samples the
timeseries had (to 5-7% for millions of samples). This I can't fully explain,
but perhaps garbage collection issues were involved.
*WHEN TO USE THE NEW TIMESTAMP TYPE*
The new clientmodel.Timestamp type should be used whenever time
calculations are either directly or indirectly related to sample
timestamps.
For example:
- the timestamp of a sample itself
- all kinds of watermarks
- anything that may become or is compared to a sample timestamp (like the timestamp
passed into Target.Scrape()).
When to still use time.Time:
- for measuring durations/times not related to sample timestamps, like duration
telemetry exporting, timers that indicate how frequently to execute some
action, etc.
*NOTE ON OPERATOR OPTIMIZATION TESTS*
We don't use operator optimization code anymore, but it still lives in
the code as dead code. It still has tests, but I couldn't get all of them to
pass with the new timestamp format. I commented out the failing cases for now,
but we should probably remove the dead code soon. I just didn't want to do that
in the same change as this.
Change-Id: I821787414b0debe85c9fffaeb57abd453727af0f
2013-10-28 06:35:02 -07:00
|
|
|
time: clientmodel.TimestampFromUnix(100),
|
2013-02-08 09:03:26 -08:00
|
|
|
},
|
|
|
|
},
|
|
|
|
},
|
|
|
|
out: out{
|
|
|
|
{
|
|
|
|
fingerprint: "00000000000000000000-a-4-a",
|
|
|
|
},
|
|
|
|
{
|
|
|
|
fingerprint: "00000000000000001111-a-4-a",
|
|
|
|
},
|
|
|
|
},
|
|
|
|
},
|
|
|
|
// // Ensure that the fingerprint-timestamp pairs are sorted in proper order.
|
|
|
|
{
|
|
|
|
in: in{
|
|
|
|
atTimes: []atTime{
|
|
|
|
{
|
|
|
|
fingerprint: "1111-a-4-a",
|
Use custom timestamp type for sample timestamps and related code.
So far we've been using Go's native time.Time for anything related to sample
timestamps. Since the range of time.Time is much bigger than what we need, this
has created two problems:
- there could be time.Time values which were out of the range/precision of the
time type that we persist to disk, therefore causing incorrectly ordered keys.
One bug caused by this was:
https://github.com/prometheus/prometheus/issues/367
It would be good to use a timestamp type that's more closely aligned with
what the underlying storage supports.
- sizeof(time.Time) is 192, while Prometheus should be ok with a single 64-bit
Unix timestamp (possibly even a 32-bit one). Since we store samples in large
numbers, this seriously affects memory usage. Furthermore, copying/working
with the data will be faster if it's smaller.
*MEMORY USAGE RESULTS*
Initial memory usage comparisons for a running Prometheus with 1 timeseries and
100,000 samples show roughly a 13% decrease in total (VIRT) memory usage. In my
tests, this advantage for some reason decreased a bit the more samples the
timeseries had (to 5-7% for millions of samples). This I can't fully explain,
but perhaps garbage collection issues were involved.
*WHEN TO USE THE NEW TIMESTAMP TYPE*
The new clientmodel.Timestamp type should be used whenever time
calculations are either directly or indirectly related to sample
timestamps.
For example:
- the timestamp of a sample itself
- all kinds of watermarks
- anything that may become or is compared to a sample timestamp (like the timestamp
passed into Target.Scrape()).
When to still use time.Time:
- for measuring durations/times not related to sample timestamps, like duration
telemetry exporting, timers that indicate how frequently to execute some
action, etc.
*NOTE ON OPERATOR OPTIMIZATION TESTS*
We don't use operator optimization code anymore, but it still lives in
the code as dead code. It still has tests, but I couldn't get all of them to
pass with the new timestamp format. I commented out the failing cases for now,
but we should probably remove the dead code soon. I just didn't want to do that
in the same change as this.
Change-Id: I821787414b0debe85c9fffaeb57abd453727af0f
2013-10-28 06:35:02 -07:00
|
|
|
time: clientmodel.TimestampFromUnix(100),
|
2013-02-08 09:03:26 -08:00
|
|
|
},
|
|
|
|
{
|
|
|
|
fingerprint: "1111-a-4-a",
|
Use custom timestamp type for sample timestamps and related code.
So far we've been using Go's native time.Time for anything related to sample
timestamps. Since the range of time.Time is much bigger than what we need, this
has created two problems:
- there could be time.Time values which were out of the range/precision of the
time type that we persist to disk, therefore causing incorrectly ordered keys.
One bug caused by this was:
https://github.com/prometheus/prometheus/issues/367
It would be good to use a timestamp type that's more closely aligned with
what the underlying storage supports.
- sizeof(time.Time) is 192, while Prometheus should be ok with a single 64-bit
Unix timestamp (possibly even a 32-bit one). Since we store samples in large
numbers, this seriously affects memory usage. Furthermore, copying/working
with the data will be faster if it's smaller.
*MEMORY USAGE RESULTS*
Initial memory usage comparisons for a running Prometheus with 1 timeseries and
100,000 samples show roughly a 13% decrease in total (VIRT) memory usage. In my
tests, this advantage for some reason decreased a bit the more samples the
timeseries had (to 5-7% for millions of samples). This I can't fully explain,
but perhaps garbage collection issues were involved.
*WHEN TO USE THE NEW TIMESTAMP TYPE*
The new clientmodel.Timestamp type should be used whenever time
calculations are either directly or indirectly related to sample
timestamps.
For example:
- the timestamp of a sample itself
- all kinds of watermarks
- anything that may become or is compared to a sample timestamp (like the timestamp
passed into Target.Scrape()).
When to still use time.Time:
- for measuring durations/times not related to sample timestamps, like duration
telemetry exporting, timers that indicate how frequently to execute some
action, etc.
*NOTE ON OPERATOR OPTIMIZATION TESTS*
We don't use operator optimization code anymore, but it still lives in
the code as dead code. It still has tests, but I couldn't get all of them to
pass with the new timestamp format. I commented out the failing cases for now,
but we should probably remove the dead code soon. I just didn't want to do that
in the same change as this.
Change-Id: I821787414b0debe85c9fffaeb57abd453727af0f
2013-10-28 06:35:02 -07:00
|
|
|
time: clientmodel.TimestampFromUnix(200),
|
2013-02-08 09:03:26 -08:00
|
|
|
},
|
|
|
|
{
|
|
|
|
fingerprint: "0-a-4-a",
|
Use custom timestamp type for sample timestamps and related code.
So far we've been using Go's native time.Time for anything related to sample
timestamps. Since the range of time.Time is much bigger than what we need, this
has created two problems:
- there could be time.Time values which were out of the range/precision of the
time type that we persist to disk, therefore causing incorrectly ordered keys.
One bug caused by this was:
https://github.com/prometheus/prometheus/issues/367
It would be good to use a timestamp type that's more closely aligned with
what the underlying storage supports.
- sizeof(time.Time) is 192, while Prometheus should be ok with a single 64-bit
Unix timestamp (possibly even a 32-bit one). Since we store samples in large
numbers, this seriously affects memory usage. Furthermore, copying/working
with the data will be faster if it's smaller.
*MEMORY USAGE RESULTS*
Initial memory usage comparisons for a running Prometheus with 1 timeseries and
100,000 samples show roughly a 13% decrease in total (VIRT) memory usage. In my
tests, this advantage for some reason decreased a bit the more samples the
timeseries had (to 5-7% for millions of samples). This I can't fully explain,
but perhaps garbage collection issues were involved.
*WHEN TO USE THE NEW TIMESTAMP TYPE*
The new clientmodel.Timestamp type should be used whenever time
calculations are either directly or indirectly related to sample
timestamps.
For example:
- the timestamp of a sample itself
- all kinds of watermarks
- anything that may become or is compared to a sample timestamp (like the timestamp
passed into Target.Scrape()).
When to still use time.Time:
- for measuring durations/times not related to sample timestamps, like duration
telemetry exporting, timers that indicate how frequently to execute some
action, etc.
*NOTE ON OPERATOR OPTIMIZATION TESTS*
We don't use operator optimization code anymore, but it still lives in
the code as dead code. It still has tests, but I couldn't get all of them to
pass with the new timestamp format. I commented out the failing cases for now,
but we should probably remove the dead code soon. I just didn't want to do that
in the same change as this.
Change-Id: I821787414b0debe85c9fffaeb57abd453727af0f
2013-10-28 06:35:02 -07:00
|
|
|
time: clientmodel.TimestampFromUnix(100),
|
2013-02-08 09:03:26 -08:00
|
|
|
},
|
|
|
|
{
|
|
|
|
fingerprint: "0-a-4-a",
|
Use custom timestamp type for sample timestamps and related code.
So far we've been using Go's native time.Time for anything related to sample
timestamps. Since the range of time.Time is much bigger than what we need, this
has created two problems:
- there could be time.Time values which were out of the range/precision of the
time type that we persist to disk, therefore causing incorrectly ordered keys.
One bug caused by this was:
https://github.com/prometheus/prometheus/issues/367
It would be good to use a timestamp type that's more closely aligned with
what the underlying storage supports.
- sizeof(time.Time) is 192, while Prometheus should be ok with a single 64-bit
Unix timestamp (possibly even a 32-bit one). Since we store samples in large
numbers, this seriously affects memory usage. Furthermore, copying/working
with the data will be faster if it's smaller.
*MEMORY USAGE RESULTS*
Initial memory usage comparisons for a running Prometheus with 1 timeseries and
100,000 samples show roughly a 13% decrease in total (VIRT) memory usage. In my
tests, this advantage for some reason decreased a bit the more samples the
timeseries had (to 5-7% for millions of samples). This I can't fully explain,
but perhaps garbage collection issues were involved.
*WHEN TO USE THE NEW TIMESTAMP TYPE*
The new clientmodel.Timestamp type should be used whenever time
calculations are either directly or indirectly related to sample
timestamps.
For example:
- the timestamp of a sample itself
- all kinds of watermarks
- anything that may become or is compared to a sample timestamp (like the timestamp
passed into Target.Scrape()).
When to still use time.Time:
- for measuring durations/times not related to sample timestamps, like duration
telemetry exporting, timers that indicate how frequently to execute some
action, etc.
*NOTE ON OPERATOR OPTIMIZATION TESTS*
We don't use operator optimization code anymore, but it still lives in
the code as dead code. It still has tests, but I couldn't get all of them to
pass with the new timestamp format. I commented out the failing cases for now,
but we should probably remove the dead code soon. I just didn't want to do that
in the same change as this.
Change-Id: I821787414b0debe85c9fffaeb57abd453727af0f
2013-10-28 06:35:02 -07:00
|
|
|
time: clientmodel.TimestampFromUnix(0),
|
2013-02-08 09:03:26 -08:00
|
|
|
},
|
|
|
|
},
|
|
|
|
},
|
|
|
|
out: out{
|
Remove the multi-op-per-fingerprint capability.
Currently, rendering a view is capable of handling multiple ops for
the same fingerprint efficiently. However, this capability requires a
lot of complexity in the code, which we are not using at all because
the way we assemble a viewRequest will never have more than one
operation per fingerprint.
This commit weeds out the said capability, along with all the code
needed for it. It is still possible to have more than one operation
for the same fingerprint, it will just be handled in a less efficient
way (as proven by the unit tests).
As a result, scanjob.go could be removed entirely.
This commit also contains a few related refactorings and removals of
dead code in operation.go, view,go, and freelist.go. Also, the
docstrings received some love.
Change-Id: I032b976e0880151c3f3fdb3234fb65e484f0e2e5
2014-02-27 11:09:00 -08:00
|
|
|
{
|
|
|
|
fingerprint: "00000000000000000000-a-4-a",
|
|
|
|
},
|
2013-02-08 09:03:26 -08:00
|
|
|
{
|
|
|
|
fingerprint: "00000000000000000000-a-4-a",
|
|
|
|
},
|
|
|
|
{
|
|
|
|
fingerprint: "00000000000000001111-a-4-a",
|
|
|
|
},
|
Remove the multi-op-per-fingerprint capability.
Currently, rendering a view is capable of handling multiple ops for
the same fingerprint efficiently. However, this capability requires a
lot of complexity in the code, which we are not using at all because
the way we assemble a viewRequest will never have more than one
operation per fingerprint.
This commit weeds out the said capability, along with all the code
needed for it. It is still possible to have more than one operation
for the same fingerprint, it will just be handled in a less efficient
way (as proven by the unit tests).
As a result, scanjob.go could be removed entirely.
This commit also contains a few related refactorings and removals of
dead code in operation.go, view,go, and freelist.go. Also, the
docstrings received some love.
Change-Id: I032b976e0880151c3f3fdb3234fb65e484f0e2e5
2014-02-27 11:09:00 -08:00
|
|
|
{
|
|
|
|
fingerprint: "00000000000000001111-a-4-a",
|
|
|
|
},
|
2013-02-08 09:03:26 -08:00
|
|
|
},
|
|
|
|
},
|
|
|
|
// Ensure grouping of operations
|
|
|
|
{
|
|
|
|
in: in{
|
|
|
|
atTimes: []atTime{
|
|
|
|
{
|
|
|
|
fingerprint: "1111-a-4-a",
|
Use custom timestamp type for sample timestamps and related code.
So far we've been using Go's native time.Time for anything related to sample
timestamps. Since the range of time.Time is much bigger than what we need, this
has created two problems:
- there could be time.Time values which were out of the range/precision of the
time type that we persist to disk, therefore causing incorrectly ordered keys.
One bug caused by this was:
https://github.com/prometheus/prometheus/issues/367
It would be good to use a timestamp type that's more closely aligned with
what the underlying storage supports.
- sizeof(time.Time) is 192, while Prometheus should be ok with a single 64-bit
Unix timestamp (possibly even a 32-bit one). Since we store samples in large
numbers, this seriously affects memory usage. Furthermore, copying/working
with the data will be faster if it's smaller.
*MEMORY USAGE RESULTS*
Initial memory usage comparisons for a running Prometheus with 1 timeseries and
100,000 samples show roughly a 13% decrease in total (VIRT) memory usage. In my
tests, this advantage for some reason decreased a bit the more samples the
timeseries had (to 5-7% for millions of samples). This I can't fully explain,
but perhaps garbage collection issues were involved.
*WHEN TO USE THE NEW TIMESTAMP TYPE*
The new clientmodel.Timestamp type should be used whenever time
calculations are either directly or indirectly related to sample
timestamps.
For example:
- the timestamp of a sample itself
- all kinds of watermarks
- anything that may become or is compared to a sample timestamp (like the timestamp
passed into Target.Scrape()).
When to still use time.Time:
- for measuring durations/times not related to sample timestamps, like duration
telemetry exporting, timers that indicate how frequently to execute some
action, etc.
*NOTE ON OPERATOR OPTIMIZATION TESTS*
We don't use operator optimization code anymore, but it still lives in
the code as dead code. It still has tests, but I couldn't get all of them to
pass with the new timestamp format. I commented out the failing cases for now,
but we should probably remove the dead code soon. I just didn't want to do that
in the same change as this.
Change-Id: I821787414b0debe85c9fffaeb57abd453727af0f
2013-10-28 06:35:02 -07:00
|
|
|
time: clientmodel.TimestampFromUnix(100),
|
2013-02-08 09:03:26 -08:00
|
|
|
},
|
|
|
|
},
|
|
|
|
atRanges: []atRange{
|
|
|
|
{
|
|
|
|
fingerprint: "1111-a-4-a",
|
Use custom timestamp type for sample timestamps and related code.
So far we've been using Go's native time.Time for anything related to sample
timestamps. Since the range of time.Time is much bigger than what we need, this
has created two problems:
- there could be time.Time values which were out of the range/precision of the
time type that we persist to disk, therefore causing incorrectly ordered keys.
One bug caused by this was:
https://github.com/prometheus/prometheus/issues/367
It would be good to use a timestamp type that's more closely aligned with
what the underlying storage supports.
- sizeof(time.Time) is 192, while Prometheus should be ok with a single 64-bit
Unix timestamp (possibly even a 32-bit one). Since we store samples in large
numbers, this seriously affects memory usage. Furthermore, copying/working
with the data will be faster if it's smaller.
*MEMORY USAGE RESULTS*
Initial memory usage comparisons for a running Prometheus with 1 timeseries and
100,000 samples show roughly a 13% decrease in total (VIRT) memory usage. In my
tests, this advantage for some reason decreased a bit the more samples the
timeseries had (to 5-7% for millions of samples). This I can't fully explain,
but perhaps garbage collection issues were involved.
*WHEN TO USE THE NEW TIMESTAMP TYPE*
The new clientmodel.Timestamp type should be used whenever time
calculations are either directly or indirectly related to sample
timestamps.
For example:
- the timestamp of a sample itself
- all kinds of watermarks
- anything that may become or is compared to a sample timestamp (like the timestamp
passed into Target.Scrape()).
When to still use time.Time:
- for measuring durations/times not related to sample timestamps, like duration
telemetry exporting, timers that indicate how frequently to execute some
action, etc.
*NOTE ON OPERATOR OPTIMIZATION TESTS*
We don't use operator optimization code anymore, but it still lives in
the code as dead code. It still has tests, but I couldn't get all of them to
pass with the new timestamp format. I commented out the failing cases for now,
but we should probably remove the dead code soon. I just didn't want to do that
in the same change as this.
Change-Id: I821787414b0debe85c9fffaeb57abd453727af0f
2013-10-28 06:35:02 -07:00
|
|
|
from: clientmodel.TimestampFromUnix(100),
|
|
|
|
through: clientmodel.TimestampFromUnix(1000),
|
2013-02-08 09:03:26 -08:00
|
|
|
},
|
|
|
|
{
|
|
|
|
fingerprint: "1111-a-4-a",
|
Use custom timestamp type for sample timestamps and related code.
So far we've been using Go's native time.Time for anything related to sample
timestamps. Since the range of time.Time is much bigger than what we need, this
has created two problems:
- there could be time.Time values which were out of the range/precision of the
time type that we persist to disk, therefore causing incorrectly ordered keys.
One bug caused by this was:
https://github.com/prometheus/prometheus/issues/367
It would be good to use a timestamp type that's more closely aligned with
what the underlying storage supports.
- sizeof(time.Time) is 192, while Prometheus should be ok with a single 64-bit
Unix timestamp (possibly even a 32-bit one). Since we store samples in large
numbers, this seriously affects memory usage. Furthermore, copying/working
with the data will be faster if it's smaller.
*MEMORY USAGE RESULTS*
Initial memory usage comparisons for a running Prometheus with 1 timeseries and
100,000 samples show roughly a 13% decrease in total (VIRT) memory usage. In my
tests, this advantage for some reason decreased a bit the more samples the
timeseries had (to 5-7% for millions of samples). This I can't fully explain,
but perhaps garbage collection issues were involved.
*WHEN TO USE THE NEW TIMESTAMP TYPE*
The new clientmodel.Timestamp type should be used whenever time
calculations are either directly or indirectly related to sample
timestamps.
For example:
- the timestamp of a sample itself
- all kinds of watermarks
- anything that may become or is compared to a sample timestamp (like the timestamp
passed into Target.Scrape()).
When to still use time.Time:
- for measuring durations/times not related to sample timestamps, like duration
telemetry exporting, timers that indicate how frequently to execute some
action, etc.
*NOTE ON OPERATOR OPTIMIZATION TESTS*
We don't use operator optimization code anymore, but it still lives in
the code as dead code. It still has tests, but I couldn't get all of them to
pass with the new timestamp format. I commented out the failing cases for now,
but we should probably remove the dead code soon. I just didn't want to do that
in the same change as this.
Change-Id: I821787414b0debe85c9fffaeb57abd453727af0f
2013-10-28 06:35:02 -07:00
|
|
|
from: clientmodel.TimestampFromUnix(100),
|
|
|
|
through: clientmodel.TimestampFromUnix(9000),
|
2013-02-08 09:03:26 -08:00
|
|
|
},
|
|
|
|
},
|
|
|
|
},
|
|
|
|
out: out{
|
|
|
|
{
|
|
|
|
fingerprint: "00000000000000001111-a-4-a",
|
|
|
|
},
|
Remove the multi-op-per-fingerprint capability.
Currently, rendering a view is capable of handling multiple ops for
the same fingerprint efficiently. However, this capability requires a
lot of complexity in the code, which we are not using at all because
the way we assemble a viewRequest will never have more than one
operation per fingerprint.
This commit weeds out the said capability, along with all the code
needed for it. It is still possible to have more than one operation
for the same fingerprint, it will just be handled in a less efficient
way (as proven by the unit tests).
As a result, scanjob.go could be removed entirely.
This commit also contains a few related refactorings and removals of
dead code in operation.go, view,go, and freelist.go. Also, the
docstrings received some love.
Change-Id: I032b976e0880151c3f3fdb3234fb65e484f0e2e5
2014-02-27 11:09:00 -08:00
|
|
|
{
|
|
|
|
fingerprint: "00000000000000001111-a-4-a",
|
|
|
|
},
|
|
|
|
{
|
|
|
|
fingerprint: "00000000000000001111-a-4-a",
|
|
|
|
},
|
2013-02-08 09:03:26 -08:00
|
|
|
},
|
|
|
|
},
|
|
|
|
}
|
|
|
|
|
|
|
|
for i, scenario := range scenarios {
|
Remove the multi-op-per-fingerprint capability.
Currently, rendering a view is capable of handling multiple ops for
the same fingerprint efficiently. However, this capability requires a
lot of complexity in the code, which we are not using at all because
the way we assemble a viewRequest will never have more than one
operation per fingerprint.
This commit weeds out the said capability, along with all the code
needed for it. It is still possible to have more than one operation
for the same fingerprint, it will just be handled in a less efficient
way (as proven by the unit tests).
As a result, scanjob.go could be removed entirely.
This commit also contains a few related refactorings and removals of
dead code in operation.go, view,go, and freelist.go. Also, the
docstrings received some love.
Change-Id: I032b976e0880151c3f3fdb3234fb65e484f0e2e5
2014-02-27 11:09:00 -08:00
|
|
|
builder := NewViewRequestBuilder()
|
2013-02-08 09:03:26 -08:00
|
|
|
|
|
|
|
for _, atTime := range scenario.in.atTimes {
|
2013-06-25 05:02:27 -07:00
|
|
|
fingerprint := &clientmodel.Fingerprint{}
|
|
|
|
fingerprint.LoadFromString(atTime.fingerprint)
|
2013-02-08 09:03:26 -08:00
|
|
|
builder.GetMetricAtTime(fingerprint, atTime.time)
|
|
|
|
}
|
|
|
|
|
|
|
|
for _, atInterval := range scenario.in.atIntervals {
|
2013-06-25 05:02:27 -07:00
|
|
|
fingerprint := &clientmodel.Fingerprint{}
|
|
|
|
fingerprint.LoadFromString(atInterval.fingerprint)
|
2013-02-08 09:03:26 -08:00
|
|
|
builder.GetMetricAtInterval(fingerprint, atInterval.from, atInterval.through, atInterval.interval)
|
|
|
|
}
|
|
|
|
|
|
|
|
for _, atRange := range scenario.in.atRanges {
|
2013-06-25 05:02:27 -07:00
|
|
|
fingerprint := &clientmodel.Fingerprint{}
|
|
|
|
fingerprint.LoadFromString(atRange.fingerprint)
|
2013-02-08 09:03:26 -08:00
|
|
|
builder.GetMetricRange(fingerprint, atRange.from, atRange.through)
|
|
|
|
}
|
|
|
|
|
|
|
|
for j, job := range scenario.out {
|
Remove the multi-op-per-fingerprint capability.
Currently, rendering a view is capable of handling multiple ops for
the same fingerprint efficiently. However, this capability requires a
lot of complexity in the code, which we are not using at all because
the way we assemble a viewRequest will never have more than one
operation per fingerprint.
This commit weeds out the said capability, along with all the code
needed for it. It is still possible to have more than one operation
for the same fingerprint, it will just be handled in a less efficient
way (as proven by the unit tests).
As a result, scanjob.go could be removed entirely.
This commit also contains a few related refactorings and removals of
dead code in operation.go, view,go, and freelist.go. Also, the
docstrings received some love.
Change-Id: I032b976e0880151c3f3fdb3234fb65e484f0e2e5
2014-02-27 11:09:00 -08:00
|
|
|
got := builder.PopOp()
|
|
|
|
if got.Fingerprint().String() != job.fingerprint {
|
|
|
|
t.Errorf("%d.%d. expected fingerprint %s, got %s", i, j, job.fingerprint, got.Fingerprint())
|
2013-02-08 09:03:26 -08:00
|
|
|
}
|
|
|
|
}
|
Remove the multi-op-per-fingerprint capability.
Currently, rendering a view is capable of handling multiple ops for
the same fingerprint efficiently. However, this capability requires a
lot of complexity in the code, which we are not using at all because
the way we assemble a viewRequest will never have more than one
operation per fingerprint.
This commit weeds out the said capability, along with all the code
needed for it. It is still possible to have more than one operation
for the same fingerprint, it will just be handled in a less efficient
way (as proven by the unit tests).
As a result, scanjob.go could be removed entirely.
This commit also contains a few related refactorings and removals of
dead code in operation.go, view,go, and freelist.go. Also, the
docstrings received some love.
Change-Id: I032b976e0880151c3f3fdb3234fb65e484f0e2e5
2014-02-27 11:09:00 -08:00
|
|
|
if builder.HasOp() {
|
|
|
|
t.Error("Expected builder to have no scan jobs left.")
|
|
|
|
}
|
2013-02-08 09:03:26 -08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
func TestBuilder(t *testing.T) {
|
|
|
|
testBuilder(t)
|
|
|
|
}
|
|
|
|
|
|
|
|
func BenchmarkBuilder(b *testing.B) {
|
|
|
|
for i := 0; i < b.N; i++ {
|
|
|
|
testBuilder(b)
|
|
|
|
}
|
|
|
|
}
|