Github user clockfly closed the pull request at:
https://github.com/apache/storm/pull/120
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/120#issuecomment-146773478
Sure.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled
Github user clockfly commented on a diff in the pull request:
https://github.com/apache/storm/pull/495#discussion_r27561927
--- Diff: storm-core/src/jvm/backtype/storm/utils/Utils.java ---
@@ -135,35 +137,68 @@ public static void sleep(long millis
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/495#issuecomment-88295467
+1
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-82196521
I will ask Ted to close this.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/463#issuecomment-78255547
+1
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/429#issuecomment-74084548
+1
On Thu, Feb 12, 2015 at 6:44 PM, Michael G. Noll
wrote:
> Thanks for your feedback, Nathan.
>
> As far as I understand this p
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/428#issuecomment-73932513
I see, there are multiple remote-merging, which then make rebase
impossible.
How about create a patch file against master and then apply the patch with
new
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/428#issuecomment-73925561
Something like git rebase -i upstream/master
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/428#issuecomment-73925451
Do we want to fix the commit log?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/428#issuecomment-73925368
+1
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-73910685
@miguno,
Go ahead and submit a new PR regarding storm-329, make sure give Ted his
credits:)
---
If your project is set up for it, you can reply to this email
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-73910187
So, I don't think your TODO comment is a issue, it is actually designed
like this, how do you think?
---
If your project is set up for it, you can reply to this
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-73909944
@miguno,
I am sure you are aware that you can still send data to this channel even
if channel.isWritable return false.
check http://netty.io/4.0/api/io
Github user clockfly closed the pull request at:
https://github.com/apache/storm/pull/45
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/45#issuecomment-73662672
it is a long time that this PR not get updated.
Let's close it for now, I prefer to create new PR if this patch is still
needed.
---
If your proje
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-73651894
+1 on the update, the patch is well written.
I made a few comments under reivew panel of
https://github.com/miguno/storm/commit
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/120#issuecomment-69435220
@revans,
Feel free to do what you want, change it, or replace it.:)
---
If your project is set up for it, you can reply to this email and have your
reply appear
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-67047625
@nathanmarz ,
I'd like to explain why I need to change worker.clj.
This was also motivated by a legacy TODO in in zmq.clj.
https://githu
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-66620748
+1
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-66620738
@tedxia
Thanks, I think this will work.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-65763659
Thanks Tedï¼
Yes, we need a fine-grained lock at âsynchronrized connect() â.
I see you changed it to use schedule Runnable, it may cause deadlock
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-63443152
It seems common in storm UT to have random failures.
We may need to clean Storm UT to make it faster and more robust.
---
If your project is set up for it, you
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/120#issuecomment-63431309
The original behavior of findAndReadConfigFile() is to locate config file
on classpath.
findResources(name) will not be empty when name exists on classpath
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/310#issuecomment-62510083
Seems this is duplicate of STORM-188
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-62509705
@tedxia
The thread pool size of clientScheduleService is decided by worker number
(also >=1 and <= 10).
For example, if there are 2 worker, the pool s
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/304#issuecomment-61921292
+1
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/286#issuecomment-61657845
+1
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/302#issuecomment-61658959
+1 for the ideas. We should have more metrics about the connection status.
From other side, current storm metric system code is too hard to read and
too hard to
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/292#issuecomment-61644279
committed to trunk.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/290#issuecomment-61644034
committed into trunk
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/290#issuecomment-61643778
+1
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/289#issuecomment-61643705
+1
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/279#issuecomment-61641290
+1
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/252#issuecomment-61639240
merged.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/252#issuecomment-61639159
+1
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/282#issuecomment-61636291
+1
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/295#issuecomment-61634029
To make the whole topology responsive, the spout need to stay active to
pull data frequently from acker or system tick.
When setting
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/305#issuecomment-61632248
dashengju is describling a use scenario that parent process blocks while
trying to read error stream data from sub-process (parent java space throw an
exception, which
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/305#issuecomment-61631943
@itaifrenkel,
```
When you say "the parent process fails" do you mean that the worker
process is no longer running?
```
No, it mean
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/303#issuecomment-61631456
@NareshKosgi,
Have you considered other special character that may used as topology name?
Like the table, #, &, (, and etc..
Is it possble that
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/305#issuecomment-61631128
@dashengju
When parent fails while child shell process is running correctly, should we
call cleanup() to kill the suprocess first?
This has two
Github user clockfly commented on a diff in the pull request:
https://github.com/apache/storm/pull/305#discussion_r19798787
--- Diff: storm-core/src/jvm/backtype/storm/utils/ShellProcess.java ---
@@ -135,7 +135,14 @@ public void logErrorStream() {
public String
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/304#issuecomment-61605948
Also,
can you remove the braket "[]" around the storm id.
github cannot link with apache jira if you have that.
---
If your project is set up f
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/304#issuecomment-61605751
@Sergeant007,
Nice find about the channel = null.
I am a little scared by the complexity of the test code,
I am +1 if you:
1. add a comment
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/298#issuecomment-61601572
@caofangkun
User can set native lib by config option:
```
confvalue("java.library.path", extrajars),
```
Should that be flexi
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/292#issuecomment-61601001
@lukedemi
Please add the JIRA number to the title of the pull request?
---
If your project is set up for it, you can reply to this email and have your
reply
GitHub user clockfly reopened a pull request:
https://github.com/apache/storm/pull/45
STORM-253: Allow storm worker to use dynamic port instead of pre-configed.
Add a ne...
PR for [STORM-253](https://issues.apache.org/jira/browse/STORM-253)
Allow supervisor to decide
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/292#issuecomment-61600425
+1
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-61600325
@nathanmarz @revans2
Do you want to take a look at it?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/252#issuecomment-61600229
@revans2, looks like you mark it as resolved. Have you merged ths into
trunk in other PR?
---
If your project is set up for it, you can reply to this email and have
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/120#issuecomment-61599764
@d2r , sorry it takes so long.
Now, the patch is synced with upstream.
---
If your project is set up for it, you can reply to this email and have your
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/257#issuecomment-61579721
@HeartSaVioR,
The performance is not a concern, since taskToQueueId will only be modified
for a few times.
+1 for the patch
---
If your project is set
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-61410622
Hi HeartSaVioR,
For acked topology, there is at least once delivery gurantee.
When a tuple is dropped, the tuple cached at spout side will timeout, and
it
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-61393817
Hi Ted,
Can you try this on your live cluster and contribute some real case test
result?
---
If your project is set up for it, you can reply to this email and
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-61393801
Summary of the test:
UT pass
STORM-404 pass
STORM-510 pass
Performacne regression pass
---
If your project is set up for it, you
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-61393103
High Availability test
===
test scenario: 4 machine A,B,C,D, 4 worker, 1 worker on each machine
test case1(STORM-404): on machine A, kill
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-61392411
When I reverted STORM-350, and test it again.
There is no more message loss.
---
If your project is set up for it, you can reply to this email and have your
reply
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-61373371
I traced back the bug of message loss.
I found this issue is introduced by storm-350.
(https://github.com/apache/storm/pull/134/files#diff
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-61287622
storm-510 is also taken care of. since we don't block the connection to
other workers.
Now, Client.send(msgs) will never block.
---
If your project is s
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-61287291
During this test, I found there was message loss, but it was not caused by
this new patch.
I traced back, seems the fault is introduced with changes between
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-61286844
About performance test:
===
I tested the performance of new patch.
It has no sigificant difference with storm-0.92
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-61071050
The UT pass.
Tests run: 174, Assertions: 101503, Failures: 0, Errors: 0
[INFO]
[INFO] Storm . SUCCESS
Github user clockfly commented on a diff in the pull request:
https://github.com/apache/storm/pull/268#discussion_r19519682
--- Diff: storm-core/src/clj/backtype/storm/daemon/worker.clj ---
@@ -378,9 +392,15 @@
_ (refresh-connections nil)
_ (refresh
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-60871537
Ted raised a good point that, ```storm.messaging.netty.max_retries```
should also be larger than worker session timeout.
Current
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-60871259
```
If worker doesn't get update of B from zookeeper after 300 reconnection
retries, should we exit the worker or let worker continues to
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-60870953
@tedxia
I got a chance to chat with Ted online. In summary, he is descrbing the
following case ï¼worker A -> worker B):
1. B dies
2. after zk sess
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-60863620
Hi Ted,
```
This will still cause another worker crash.
Now there are two things happen in parallel, first nimbus inform worker A
that worker B is not
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-59676638
The updated patch still requires more function test and performance
regression test.
---
If your project is set up for it, you can reply to this email and have your
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-59676137
```
When target worker is down, the data sending to other target worker should
not be blocked.
The approach we currently using is to drop messages when connection
Github user clockfly commented on a diff in the pull request:
https://github.com/apache/storm/pull/268#discussion_r19065812
--- Diff: storm-core/src/clj/backtype/storm/daemon/worker.clj ---
@@ -330,6 +330,20 @@
(.send drainer node+port->soc
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-59674932
To retain the comments, I modified ted's branch.
The update will tries to slove these three problems:
1. When target worker is down, the source worker s
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-59487520
Support we are sending data from worker A to worker B, to solve
STORM-404(Worker on one machine crashes due to a failure of another worker on
another machine
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-59317503
For item 3, we can move that to a sub-task.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your
Github user clockfly commented on the pull request:
https://github.com/apache/storm/pull/268#issuecomment-59317470
Hi Ted,
You are trying to achive three things:
1. When a worker is down, we don't want the upstream worker to fail in a
chaining manner.
2. When
Github user clockfly commented on the pull request:
https://github.com/apache/incubator-storm/pull/103#issuecomment-44925955
Thanks @miofthena,
fixed at 0bca173! I should have done more test on new checkin.
---
If your project is set up for it, you can reply to this email
76 matches
Mail list logo
