I can do it Joel. As Mike said, this is a bad bug that can lead to data
lost...

2017-04-20 22:43 GMT+02:00 Joel Bernstein <[email protected]>:

> I'm just uploading RC2 now. We can respin, but I won't be available for a
> respin this evening or tomorrow. If someone else wants to step in thats ok.
>
> Joel Bernstein
> http://joelsolr.blogspot.com/
>
> On Thu, Apr 20, 2017 at 4:28 PM, Michael McCandless <
> [email protected]> wrote:
>
>> Hi Joel,
>>
>> LUCENE-7791 is actually quite a bad bug, and the patch looks low risk and
>> committable; maybe we should merge it for 6.5.1?
>>
>> Mike McCandless
>>
>> http://blog.mikemccandless.com
>>
>> On Thu, Apr 20, 2017 at 3:25 PM, Steve Rowe <[email protected]> wrote:
>>
>>> Joel,
>>>
>>> I’ve created <https://issues.apache.org/jira/browse/LUCENE-7793> and <
>>> https://issues.apache.org/jira/browse/LUCENE-7794> to add precommit
>>> checking (or the equivalent) to buildAndPushRelease.py and
>>> smokeTestRelease.py, respectively.
>>>
>>> --
>>> Steve
>>> www.lucidworks.com
>>>
>>> > On Apr 20, 2017, at 10:29 AM, Steve Rowe <[email protected]> wrote:
>>> >
>>> > I was thinking that if buildAndPushRelease.py can identify problems
>>> early, the RM won’t have to later redo work when a problem is found, e.g.
>>> by the smoke tester.  IMHO the extra 10 minutes (on my laptop anyway) from
>>> auto-running precommit would be worth it.
>>> >
>>> > But I agree that the smoke tester should be heavy, since it’s possible
>>> that the RM had weird local conditions that weren’t caught by the release
>>> production tools.
>>> >
>>> > --
>>> > Steve
>>> > www.lucidworks.com
>>> >
>>> >> On Apr 20, 2017, at 10:23 AM, Joel Bernstein <[email protected]>
>>> wrote:
>>> >>
>>> >> They are both quite long processes. Perhaps we should look at
>>> eliminating some of the overlap?
>>> >>
>>> >> Since the smokeTesters is run by those voting on the release maybe
>>> this should be the heavier of the two.
>>> >>
>>> >> Joel Bernstein
>>> >> http://joelsolr.blogspot.com/
>>> >>
>>> >> On Thu, Apr 20, 2017 at 10:00 AM, Steve Rowe <[email protected]>
>>> wrote:
>>> >> Also, I think buildAndPushRelease.py should invoke “precommit" too,
>>> (near) where it runs 'ant clean test'.
>>> >>
>>> >> --
>>> >> Steve
>>> >> www.lucidworks.com
>>> >>
>>> >>> On Apr 20, 2017, at 9:51 AM, Steve Rowe <[email protected]> wrote:
>>> >>>
>>> >>> That’s bizarre.
>>> >>>
>>> >>> The smoke tester runs ‘ant validate’, which does *not* include
>>> "documentation-lint” (which includes "-ecj-javadoc-lint”, where unused
>>> imports are looked for).
>>> >>>
>>> >>> By contrast, ‘ant precommit’ runs “check-working-copy”, “validate”,
>>> and “documentation-lint”.
>>> >>>
>>> >>> Jenkins caught this not because of the smoke tester, but rather
>>> because the “nightly-smoke” task it uses runs “-jenkins-base”, which, like
>>> “precommit”, runs “check-working-copy”, “validate”, and
>>> “documentation-lint” (among others).
>>> >>>
>>> >>> I think we should change the smoke tester to look at the same things
>>> as precommit,
>>> >>>
>>> >>> --
>>> >>> Steve
>>> >>> www.lucidworks.com
>>> >>>
>>> >>>> On Apr 20, 2017, at 9:29 AM, Joel Bernstein <[email protected]>
>>> wrote:
>>> >>>>
>>> >>>> Smoke testes passed as well
>>> >>>>
>>> >>>>
>>> >>>> Joel Bernstein
>>> >>>> http://joelsolr.blogspot.com/
>>> >>>>
>>> >>>> On Thu, Apr 20, 2017 at 9:21 AM, Steve Rowe <[email protected]>
>>> wrote:
>>> >>>> I guess precommit is only run as part of the smoke tester.
>>> >>>>
>>> >>>>> On Apr 20, 2017, at 9:20 AM, Joel Bernstein <[email protected]>
>>> wrote:
>>> >>>>>
>>> >>>>> Interesting the entire release process completed without
>>> complaining.
>>> >>>>>
>>> >>>>> Joel Bernstein
>>> >>>>> http://joelsolr.blogspot.com/
>>> >>>>>
>>> >>>>> On Thu, Apr 20, 2017 at 9:10 AM, Steve Rowe <[email protected]>
>>> wrote:
>>> >>>>> The unused import was the only problem; I’ve removed it on
>>> branch_6_5.
>>> >>>>>
>>> >>>>> Sorry, hope I didn’t cause a delay in the release process.
>>> >>>>>
>>> >>>>> --
>>> >>>>> Steve
>>> >>>>> www.lucidworks.com
>>> >>>>>
>>> >>>>>> On Apr 20, 2017, at 8:57 AM, Steve Rowe <[email protected]> wrote:
>>> >>>>>>
>>> >>>>>> Crap, Jenkins found an unused import in my commit on branch_6_5
>>> :(.  I forgot to run precommit.  Doing so now.
>>> >>>>>>
>>> >>>>>> --
>>> >>>>>> Steve
>>> >>>>>> www.lucidworks.com
>>> >>>>>>
>>> >>>>>>> On Apr 19, 2017, at 8:07 PM, Steve Rowe <[email protected]>
>>> wrote:
>>> >>>>>>>
>>> >>>>>>> Done.
>>> >>>>>>>
>>> >>>>>>> --
>>> >>>>>>> Steve
>>> >>>>>>> www.lucidworks.com
>>> >>>>>>>
>>> >>>>>>>> On Apr 19, 2017, at 7:49 PM, Steve Rowe <[email protected]>
>>> wrote:
>>> >>>>>>>>
>>> >>>>>>>> Woohoo!  Thanks Joel.  Running all tests on branch_6_5 now.
>>> >>>>>>>>
>>> >>>>>>>> --
>>> >>>>>>>> Steve
>>> >>>>>>>> www.lucidworks.com
>>> >>>>>>>>
>>> >>>>>>>>> On Apr 19, 2017, at 7:20 PM, Joel Bernstein <
>>> [email protected]> wrote:
>>> >>>>>>>>>
>>> >>>>>>>>> I'm OK with restarting RC2. Let me know when you finish the
>>> back port.
>>> >>>>>>>>>
>>> >>>>>>>>> Joel Bernstein
>>> >>>>>>>>> http://joelsolr.blogspot.com/
>>> >>>>>>>>>
>>> >>>>>>>>> On Wed, Apr 19, 2017 at 7:11 PM, Steve Rowe <[email protected]>
>>> wrote:
>>> >>>>>>>>> Joel,
>>> >>>>>>>>>
>>> >>>>>>>>> I know you’re in-process with RC2, but if you have to restart
>>> for some reason, I’d like to include SOLR-10527.
>>> >>>>>>>>>
>>> >>>>>>>>> --
>>> >>>>>>>>> Steve
>>> >>>>>>>>> www.lucidworks.com
>>> >>>>>>>>>
>>> >>>>>>>>>> On Apr 19, 2017, at 6:57 PM, Joel Bernstein <
>>> [email protected]> wrote:
>>> >>>>>>>>>>
>>> >>>>>>>>>> I also had to change the smoke tester because of the SSL
>>> issue. Which in theory means everyone else will have to as well. I'm
>>> pushing out my changes to branch_6_5, so people can just update their
>>> branch before running the smoke tester.
>>> >>>>>>>>>>
>>> >>>>>>>>>> Joel Bernstein
>>> >>>>>>>>>> http://joelsolr.blogspot.com/
>>> >>>>>>>>>>
>>> >>>>>>>>>> On Wed, Apr 19, 2017 at 5:02 PM, Joel Bernstein <
>>> [email protected]> wrote:
>>> >>>>>>>>>> Ok, I had to push out three changes to the branch but I
>>> seemed to have moved passed this issue.
>>> >>>>>>>>>>
>>> >>>>>>>>>> Joel Bernstein
>>> >>>>>>>>>> http://joelsolr.blogspot.com/
>>> >>>>>>>>>>
>>> >>>>>>>>>> On Wed, Apr 19, 2017 at 4:50 PM, Joel Bernstein <
>>> [email protected]> wrote:
>>> >>>>>>>>>> I'll push out the changes and then revert. Hopefully the
>>> changes will do the trick after they've been pushed.
>>> >>>>>>>>>>
>>> >>>>>>>>>> Joel Bernstein
>>> >>>>>>>>>> http://joelsolr.blogspot.com/
>>> >>>>>>>>>>
>>> >>>>>>>>>> On Wed, Apr 19, 2017 at 4:37 PM, Steve Rowe <[email protected]>
>>> wrote:
>>> >>>>>>>>>> Yeah, that is a sucky part of this: you want to make
>>> modifications on the release branch, but you can’t test unless you commit
>>> (and push).
>>> >>>>>>>>>>
>>> >>>>>>>>>> Another alternative: make modifications in another checkout,
>>> then invoke the script with from an unmodified source tree.
>>> >>>>>>>>>>
>>> >>>>>>>>>> e.g.:
>>> >>>>>>>>>>
>>> >>>>>>>>>> cd ..
>>> >>>>>>>>>> cp -r lucene-solr lucene-solr-2 # assuming lucene-solr is
>>> your checkout dir
>>> >>>>>>>>>> # edit lucene-solr-2/dev-tools/scripts/buildAndPushRelease.py
>>> >>>>>>>>>> cd lucene-solr
>>> >>>>>>>>>> python3 -u 
>>> >>>>>>>>>> ../lucene-solr-2/dev-tools/scripts/buildAndPushRelease.py
>>> …
>>> >>>>>>>>>>
>>> >>>>>>>>>> (a shorter form of that is to make a copy the script alone
>>> and invoke the modified version from your checkout)
>>> >>>>>>>>>>
>>> >>>>>>>>>> --
>>> >>>>>>>>>> Steve
>>> >>>>>>>>>> www.lucidworks.com
>>> >>>>>>>>>>
>>> >>>>>>>>>>> On Apr 19, 2017, at 4:32 PM, Joel Bernstein <
>>> [email protected]> wrote:
>>> >>>>>>>>>>>
>>> >>>>>>>>>>> Ha, that doesn't work either. Now I get this error:
>>> >>>>>>>>>>>
>>> >>>>>>>>>>> RuntimeError: There are unpushed commits - "git log
>>> origin/branch_6_5.." output is:
>>> >>>>>>>>>>>
>>> >>>>>>>>>>> So, I'm going to have to push this out.
>>> >>>>>>>>>>>
>>> >>>>>>>>>>> Let's first decide if this makes sense as the way forward.
>>> >>>>>>>>>>>
>>> >>>>>>>>>>>
>>> >>>>>>>>>>>
>>> >>>>>>>>>>>
>>> >>>>>>>>>>>
>>> >>>>>>>>>>> Joel Bernstein
>>> >>>>>>>>>>> http://joelsolr.blogspot.com/
>>> >>>>>>>>>>>
>>> >>>>>>>>>>> On Wed, Apr 19, 2017 at 4:28 PM, Joel Bernstein <
>>> [email protected]> wrote:
>>> >>>>>>>>>>> Steve, when I change the script to turn off verification I
>>> get the following error:
>>> >>>>>>>>>>> RuntimeError: git clone is dirty:
>>> >>>>>>>>>>>
>>> >>>>>>>>>>> So as part of the work around I think I'm going to have to
>>> commit locally and then revert locally. Does that make sense to you as the
>>> way forward?
>>> >>>>>>>>>>>
>>> >>>>>>>>>>>
>>> >>>>>>>>>>>
>>> >>>>>>>>>>>
>>> >>>>>>>>>>>
>>> >>>>>>>>>>>
>>> >>>>>>>>>>> Joel Bernstein
>>> >>>>>>>>>>> http://joelsolr.blogspot.com/
>>> >>>>>>>>>>>
>>> >>>>>>>>>>> On Wed, Apr 19, 2017 at 4:24 PM, Alexandre Rafalovitch <
>>> [email protected]> wrote:
>>> >>>>>>>>>>> Could it be anything to do with the fact that the signer is
>>> Semantic
>>> >>>>>>>>>>> and they are being actively distrusted by - at least -
>>> Google:
>>> >>>>>>>>>>> http://www.securityweek.com/google-stops-trusting-symantec-i
>>> ssued-certificates
>>> >>>>>>>>>>>
>>> >>>>>>>>>>> Regards,
>>> >>>>>>>>>>> Alex.
>>> >>>>>>>>>>> ----
>>> >>>>>>>>>>> http://www.solr-start.com/ - Resources for Solr users, new
>>> and experienced
>>> >>>>>>>>>>>
>>> >>>>>>>>>>>
>>> >>>>>>>>>>> On 19 April 2017 at 16:22, Joel Bernstein <
>>> [email protected]> wrote:
>>> >>>>>>>>>>>> Ok, I'll turn off the cert verification. I wasn't sure if
>>> cert verification
>>> >>>>>>>>>>>> was something that was integral to the process.
>>> >>>>>>>>>>>>
>>> >>>>>>>>>>>> Joel Bernstein
>>> >>>>>>>>>>>> http://joelsolr.blogspot.com/
>>> >>>>>>>>>>>>
>>> >>>>>>>>>>>> On Wed, Apr 19, 2017 at 4:18 PM, Steve Rowe <
>>> [email protected]> wrote:
>>> >>>>>>>>>>>>>
>>> >>>>>>>>>>>>> Hi Joel,
>>> >>>>>>>>>>>>>
>>> >>>>>>>>>>>>> Not sure why this is suddenly an issue - I guess “SNI”
>>> >>>>>>>>>>>>> <https://en.wikipedia.org/wiki/Server_Name_Indication>
>>> has been enabled on
>>> >>>>>>>>>>>>> archive.apache.org?
>>> >>>>>>>>>>>>>
>>> >>>>>>>>>>>>> Some useful info here (about a “requests” lib, AFAICT an
>>> alternative to
>>> >>>>>>>>>>>>> urllib):
>>> >>>>>>>>>>>>> <http://docs.python-requests.o
>>> rg/en/master/community/faq/#what-are-hostname-doesn-t-match-errors>
>>> >>>>>>>>>>>>>
>>> >>>>>>>>>>>>> Short term, you could turn off certificate verification.
>>> The bottom
>>> >>>>>>>>>>>>> answer here uses the same lib as the script (urllib
>>> instead of urllib2,
>>> >>>>>>>>>>>>> which is assumed in the other answers on the page), to
>>> turn off certificate
>>> >>>>>>>>>>>>> verification:
>>> >>>>>>>>>>>>>
>>> >>>>>>>>>>>>>
>>> >>>>>>>>>>>>> <http://stackoverflow.com/ques
>>> tions/19268548/python-ignore-certicate-validation-urllib2>
>>> >>>>>>>>>>>>>
>>> >>>>>>>>>>>>> --
>>> >>>>>>>>>>>>> Steve
>>> >>>>>>>>>>>>> www.lucidworks.com
>>> >>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> On Apr 19, 2017, at 3:43 PM, Joel Bernstein <
>>> [email protected]> wrote:
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> It appears that it's failing on verifying the SSL cert for
>>> >>>>>>>>>>>>>> https://archive.apache.org/dist/lucene/java/.
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> Joel Bernstein
>>> >>>>>>>>>>>>>> http://joelsolr.blogspot.com/
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> On Wed, Apr 19, 2017 at 3:36 PM, Joel Bernstein <
>>> [email protected]>
>>> >>>>>>>>>>>>>> wrote:
>>> >>>>>>>>>>>>>> I started working on 6.5.1 RC2. I ran the following:
>>> >>>>>>>>>>>>>> python3 -u dev-tools/scripts/buildAndPushRelease.py
>>> --push-local
>>> >>>>>>>>>>>>>> /tmp/releases/6.5.1 --rc-num 2 --sign EE64CB1E
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> I believe this is same basic command I used for RC1. But
>>> this time I got
>>> >>>>>>>>>>>>>> a new error. At first I thought it might be incorrect
>>> keystore password but
>>> >>>>>>>>>>>>>> I changed it and it is correct. Any thoughts what the
>>> issue is?
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> Traceback (most recent call last):
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> File "dev-tools/scripts/buildAndPushRelease.py", line
>>> 313, in <module>
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> main()
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> File "dev-tools/scripts/buildAndPushRelease.py", line
>>> 294, in main
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> rev = prepare(c.root, c.version, c.key_id, c.key_password)
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> File "dev-tools/scripts/buildAndPushRelease.py", line
>>> 98, in prepare
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> checkDOAPfiles(version)
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> File "dev-tools/scripts/buildAndPushRelease.py", line
>>> 143, in
>>> >>>>>>>>>>>>>> checkDOAPfiles
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> distpage = load(url)
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> File "dev-tools/scripts/buildAndPushRelease.py", line
>>> 67, in load
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> content = urllib.request.urlopen(urlStri
>>> ng).read().decode('utf-8')
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> File
>>> >>>>>>>>>>>>>> "/Library/Frameworks/Python.fr
>>> amework/Versions/3.6/lib/python3.6/urllib/request.py",
>>> >>>>>>>>>>>>>> line 223, in urlopen
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> return opener.open(url, data, timeout)
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> File
>>> >>>>>>>>>>>>>> "/Library/Frameworks/Python.fr
>>> amework/Versions/3.6/lib/python3.6/urllib/request.py",
>>> >>>>>>>>>>>>>> line 526, in open
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> response = self._open(req, data)
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> File
>>> >>>>>>>>>>>>>> "/Library/Frameworks/Python.fr
>>> amework/Versions/3.6/lib/python3.6/urllib/request.py",
>>> >>>>>>>>>>>>>> line 544, in _open
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> '_open', req)
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> File
>>> >>>>>>>>>>>>>> "/Library/Frameworks/Python.fr
>>> amework/Versions/3.6/lib/python3.6/urllib/request.py",
>>> >>>>>>>>>>>>>> line 504, in _call_chain
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> result = func(*args)
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> File
>>> >>>>>>>>>>>>>> "/Library/Frameworks/Python.fr
>>> amework/Versions/3.6/lib/python3.6/urllib/request.py",
>>> >>>>>>>>>>>>>> line 1361, in https_open
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> context=self._context, check_hostname=self._check_hos
>>> tname)
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> File
>>> >>>>>>>>>>>>>> "/Library/Frameworks/Python.fr
>>> amework/Versions/3.6/lib/python3.6/urllib/request.py",
>>> >>>>>>>>>>>>>> line 1320, in do_open
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> raise URLError(err)
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> urllib.error.URLError: <urlopen error [SSL:
>>> CERTIFICATE_VERIFY_FAILED]
>>> >>>>>>>>>>>>>> certificate verify failed (_ssl.c:749)>
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>> Joel Bernstein
>>> >>>>>>>>>>>>>> http://joelsolr.blogspot.com/
>>> >>>>>>>>>>>>>>
>>> >>>>>>>>>>>>>
>>> >>>>>>>>>>>>>
>>> >>>>>>>>>>>>> ------------------------------
>>> ---------------------------------------
>>> >>>>>>>>>>>>> To unsubscribe, e-mail: [email protected]
>>> >>>>>>>>>>>>> For additional commands, e-mail:
>>> [email protected]
>>> >>>>>>>>>>>>>
>>> >>>>>>>>>>>>
>>> >>>>>>>>>>>
>>> >>>>>>>>>>> ------------------------------------------------------------
>>> ---------
>>> >>>>>>>>>>> To unsubscribe, e-mail: [email protected]
>>> >>>>>>>>>>> For additional commands, e-mail: [email protected]
>>> >>>>>>>>>>>
>>> >>>>>>>>>>>
>>> >>>>>>>>>>>
>>> >>>>>>>>>>
>>> >>>>>>>>>>
>>> >>>>>>>>>> ------------------------------------------------------------
>>> ---------
>>> >>>>>>>>>> To unsubscribe, e-mail: [email protected]
>>> >>>>>>>>>> For additional commands, e-mail: [email protected]
>>> >>>>>>>>>>
>>> >>>>>>>>>>
>>> >>>>>>>>>>
>>> >>>>>>>>>>
>>> >>>>>>>>>
>>> >>>>>>>>>
>>> >>>>>>>>> ------------------------------------------------------------
>>> ---------
>>> >>>>>>>>> To unsubscribe, e-mail: [email protected]
>>> >>>>>>>>> For additional commands, e-mail: [email protected]
>>> >>>>>>>>>
>>> >>>>>>>>>
>>> >>>>>>>>
>>> >>>>>>>
>>> >>>>>>
>>> >>>>>
>>> >>>>>
>>> >>>>> ------------------------------------------------------------
>>> ---------
>>> >>>>> To unsubscribe, e-mail: [email protected]
>>> >>>>> For additional commands, e-mail: [email protected]
>>> >>>>>
>>> >>>>>
>>> >>>>
>>> >>>>
>>> >>>> ------------------------------------------------------------
>>> ---------
>>> >>>> To unsubscribe, e-mail: [email protected]
>>> >>>> For additional commands, e-mail: [email protected]
>>> >>>>
>>> >>>>
>>> >>>
>>> >>
>>> >>
>>> >> ---------------------------------------------------------------------
>>> >> To unsubscribe, e-mail: [email protected]
>>> >> For additional commands, e-mail: [email protected]
>>> >>
>>> >>
>>> >
>>>
>>>
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail: [email protected]
>>> For additional commands, e-mail: [email protected]
>>>
>>>
>>
>

Reply via email to