X-CGP-ClamAV-Result: CLEAN X-VirusScanner: Niversoft's CGPClamav Helper v1.25a (ClamAV 0.103.6/26723) X-Junk-Score: 0 [] X-KAS-Score: 0 [] Return-Path: Received: from mail.hrdcstl.com ([159.203.70.82] verified) by post.selbstdenker.com (CommuniGate Pro SMTP 6.3.14) with ESMTPS id 28941306 for webobjects-dev@wocommunity.org; Thu, 17 Nov 2022 17:24:53 +0100 Received-SPF: pass receiver=post.selbstdenker.com; client-ip=159.203.70.82; envelope-from=wo@hrdcstl.com Received: from authenticated-user (mail.hrdcstl.com [159.203.70.82]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.hrdcstl.com (Postfix) with ESMTPSA id B719814ABC1 for ; Thu, 17 Nov 2022 11:24:30 -0500 (EST) DKIM-Signature: v=1; a=rsa-sha256; c=simple/simple; d=hrdcstl.com; s=mail; t=1668702270; bh=oqshPrYkh0m45HIG4cOmP9Pua88v0ESkOFe2NRTX3qo=; h=From:Subject:Date:References:To:In-Reply-To:From; b=JbG4QCT2aUwNkLIkLyWf0bSFXaTL2VTnTSDev354XQFFHf+srG6rIisGUtB4YSYtp YIYOoAGHDARA546KY11oI8DGVYaqgdVFAWQBAfeYv8DDYdcAr4CZQJ/sfzGRLSWT22 CK6mMr/OATEWxQ1UOPa2FmeOfe6GzvCtu+Bg3xeUmN/bCulCIjtgWPMXckxAscYI+S HWylw26UshP0Ib3gxyPK0+Bi9l3dgV9Kjm+65UqJk/e4ZI4zEVG91Nf41rUT57BPCX H/ALaBMVfav2NMKFF03tO9lDFLC3iTmRyLcBiuDnUlYxmvGgvogELRsGI7BNYtlyjx YfOOPXtgHF0Fw== From: Asa Steady Hardcastle Content-Type: multipart/alternative; boundary="Apple-Mail=_6F4994B0-B60F-409B-AFAF-562024896A34" Mime-Version: 1.0 Subject: Re: [WO-DEV] Non instance found but Monitor shows it as being up and running Date: Thu, 17 Nov 2022 11:24:28 -0500 References: To: WebObjects & WOnder Development In-Reply-To: Message-Id: --Apple-Mail=_6F4994B0-B60F-409B-AFAF-562024896A34 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=utf-8 Hi Markus, It looks like there are remote control REST routines for JavaMonitor. I = have never used them, but then, I didn't know they existed! =20 = https://wiki.wocommunity.org/display/documentation/Wonder+JavaMonitor+and+= wotaskd = Do you see any memory errors in the logs? Years ago we had an = application that would very occasionally hang and report "no instance = available". The WOMonitor would intermittently show that it was = running, but it would not respond to stop. At the time we would need to = kill the process (finding it with lsof filtering for the specific port) = and then launch it again. Kind of awful, and certainly not good = practice.=20 I look forward to hearing what you end up doing! ttyl, Asa > On Nov 17, 2022, at 6:45 AM, Markus Ruggiero (rucotec) = wrote: >=20 > No one? >=20 > Is there documentation for REST calls to wotaskd and/or JavaMonitor? I = haven=E2=80=99t found anything, might have been looking in the wrong = place =F0=9F=99=83 >=20 > ---markus--- >=20 >> On 15 Nov 2022, at 16:20, Markus Ruggiero (rucotec) = wrote: >>=20 >> This is a weird one. >>=20 >> We have a WOApp deployed on Ubuntu 20.04.5 LTS on a cloud server. App = is classically controlled by JavaMonitor and instance (currently only = one) is scheduled for restart every day at 5am. Once in a while the = first user in the morning is greeted by =E2=80=9Cno instance found=E2=80=9D= , but checking with JavaMonitor everything looks good and there are not = deaths reported! Manually restarting the instance clears things up. It = seems that sometimes the instance does not come up fully or the = connection with WOAdaptor to Apache breaks. Anyone having had similar = issues? We are running a fairly up-to-date Apache (Server version: = Apache/2.4.41 (Ubuntu) Server built: 2022-06-14T13:30:55). WOAdaptor = is the one built by Maik (thanks a lot for this). >>=20 >> Another weird thing is that we have various other applications = running on the same and on other similarly set up servers, none of them = having any issues. Any idea on that? >>=20 >> Is there a way to externally trigger Monitor or WOtaskd to restart an = instance? Does the REST interface support this and how would I call it = with e.g. curl? I could then create my own heartbeat script testing and = restarting instances through Apache. >>=20 >> Thanks for any help >> ---markus--- >>=20 >=20 >=20 --Apple-Mail=_6F4994B0-B60F-409B-AFAF-562024896A34 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=utf-8
Hi Markus,

It looks like there are remote control REST routines for = JavaMonitor. I have never used them, but then, I didn't know they = existed!  

https://wiki.wocommunity.org/display/documentation/Wonder+JavaM= onitor+and+wotaskd

Do you see any memory errors in the logs? Years ago we had an = application that would very occasionally hang and report "no instance = available".  The WOMonitor would intermittently show that it was = running, but it would not respond to stop. At the time we would need to = kill the process (finding it with lsof filtering for the specific port) = and then launch it again. Kind of awful, and certainly not good = practice. 

I look forward to hearing what you end up doing!

ttyl,

Asa



On Nov = 17, 2022, at 6:45 AM, Markus Ruggiero (rucotec) <markus.ruggiero@rucotec.ch> wrote:

No = one?

Is there documentation for REST calls = to wotaskd and/or JavaMonitor? I haven=E2=80=99t found anything, might = have been looking in the wrong place =F0=9F=99=83

---markus---

On 15 Nov 2022, at 16:20, Markus Ruggiero = (rucotec) <markus.ruggiero@rucotec.ch> wrote:

This is a weird one.

We have a = WOApp deployed on Ubuntu 20.04.5 LTS on a cloud server. App is = classically controlled by JavaMonitor and instance (currently only one) = is scheduled for restart every day at 5am. Once in a while the first = user in the morning is greeted by =E2=80=9Cno instance found=E2=80=9D, = but checking with JavaMonitor everything looks good and there are not = deaths reported! Manually restarting the instance clears things up. It = seems that sometimes the instance does not come up fully or the = connection with WOAdaptor to Apache breaks. Anyone having had similar = issues? We are running a fairly up-to-date Apache (Server version: = Apache/2.4.41 (Ubuntu) Server built:   2022-06-14T13:30:55). = WOAdaptor is the one built by Maik (thanks a lot for this).

Another weird thing is that we have various = other applications running on the same and on other similarly set up = servers, none of them having any issues. Any idea on that?

Is there a way to externally trigger Monitor = or WOtaskd to restart an instance? Does the REST interface support this = and how would I call it with e.g. curl? I could then create my own = heartbeat script testing and restarting instances through Apache.

Thanks for any help
---markus---




= --Apple-Mail=_6F4994B0-B60F-409B-AFAF-562024896A34--