Archives for: June 2013

Pages: 1

06/30/13

  10:42:00 am, by The Dreamer   , 902 words  
Categories: FreeBSD

Catching up on port update backlog....

This weekend, I decided it was time that I checked on port updates in my /compat/i386 FreeBSD 'system'. Which primarily exists to provide me some ports that don't build on 64-bit, namely emulators/wine-devel and net/nxserver. Don't recall the last time I used nx since I got it working, probably should check to see whether it is still working or not (probably okay on my home system, but might be broke on work one....and might see about setting it up on other work computer too).

Hmmm, hadn't updated ports since May 5th. Start with working through /usr/ports/UPDATING, run into a problem that on 20130609: AFFECTS: users of audio/flac and any port that depends on it, in that there it thinks perl depends on it (kind of an annoyance I have with dependencies....there can be miles of separation between one port and another port, but everything get's marked as depending on that very bottom port, when it in fact didn't or doesn't... Was annoying in trying to figure out why a port was marked BROKEN / DEPRECATED and not get any attention except that people should stop using it...when 100's of ports on my system depend on it. When it turns out that its one or two ports, had an option set that caused it to depend on it. While the other ports generally don't care what options are enabled in that port, just that the command exists for it...or other reason. Though there are some ports that do care about what options were used, which I had ranted about earlier...and I ran into Thunderbird also having that dependency, resulting in this kluge patch:

Code

--- Makefile.orig  2013-06-26 06:01:34.000000000 -0500
+++ Makefile  2013-06-27 20:07:04.142845537 -0500
@@ -98,6 +98,8 @@
.endif
post-patch:
+  @${REINPLACE_CMD} -e '/with SQLITE_SECURE_DELETE/s/_ERROR/_WARN/' \
+    ${WRKSRC}/mozilla/configure.in
  @${REINPLACE_CMD} -e 's|%%LOCALBASE%%|${LOCALBASE}|g' \
    ${WRKSRC}/mail/app/nsMailApp.cpp
.if ${PORT_OPTIONS:MENIGMAIL}

But, I let the portmaster -r flac run aways, with the suspicion that it would break later because perl modules that depend on perl (and not flac) wouldn't get picked up as needed to be re-installed or upgraded, due to 20130612: AFFECTS: users of lang/perl* and any port that depends on it. But, would break the re-install or upgrade of a port somewhere and abort. Which is what I found when I checked on it this morning.

So, I did a portmaster -R -r perl, and noticed that it seemed to include most of the ports that the previous portmaster hadn't done. In fact it included all of them. I also peeked in /usr/local/lib/perl5/5.14.2 and /usr/local/lib/perl5/site_perl/5.14.2 to see what perl modules had gotten missed....mainly the p5-XML-* ones that caused the previous portmaster to abort.

Though I probably should've looked to see if the second portmaster was going to address those, instead of doing them while it was asking to proceed. Because that caused it to abort when re-installing those perl modules (that I had done while it was waiting), but restarting it got things done.

That leaves the latest entry 20130627: AFFECTS: users of ports-mgmt/portmaster, which is just informational and not currently applicable.

Before running in to the flac entry, there had been "20130527: AFFECTS: users of lang/ruby18" which was pretty straight forward, since it only exists as a dependency to ports-mgmt/portupgrade, which I seldom use now...but I have other scripts that use binaries that come as part of it (namely portsclean), which I could probably replace with the portmaster way or something else. But, its not really a priority, plus who knows if I won't decide to go back to using portupgrade...which has options in its pkgtools.conf that I haven't found equivalents for with portmaster, though isn't currently an issue right now. Except perhaps that I'm holding back on updating to the latest emulators/virtualbox-ose, since I've gotten warnings from various sources to stay away from it.

The other big one is what's the portmaster equivalent to portupgrade's ALT_PKGDEP?

Will eventually run into a port that has a line one of these:

Code

RUN_DEPENDS+=   mysql-server>=0:${PORTSDIR}/databases/mysql${MYSQL_VER}-server
RUN_DEPENDS+=   mysql-server>=0:${PORTSDIR}/databases/mysql${MYSQL_VER}-server  
RUN_DEPENDS+=   mysqld_safe:${PORTSDIR}/databases/mysql55-server
RUN_DEPENDS+=   ${LOCALBASE}/bin/mysqld_safe:${PORTSDIR}/databases/mysql55-server
RUN_DEPENDS=    ${LOCALBASE}/libexec/mysqld:${PORTSDIR}/databases/mysql${MYSQL_VER}-server
RUN_DEPENDS+=   ${LOCALBASE}/libexec/mysqld:${PORTSDIR}/databases/mysql${MYSQL_VER}-server
RUN_DEPENDS+=   ${LOCALBASE}/libexec/mysqld:${PORTSDIR}/databases/mysql${MYSQL_VER}-server
RUN_DEPENDS+=   mysql-server>=0:${PORTSDIR}/databases/mysql${WANT_MYSQL_VER}-server
RUN_DEPENDS+=   mysql-server>=0:${PORTSDIR}/databases/mysql${MYSQL_VER}-server
RUN_DEPENDS+=   mysql-server>=0:${PORTSDIR}/databases/mysql${WANT_MYSQL_VER}-server
LIB_DEPENDS+=   mysqlclient:${PORTSDIR}/databases/mysql55-client
RUN_DEPENDS+=   mysqld_safe:${PORTSDIR}/databases/mysql${MYSQL_VER}-server

Where I'm using databases/percona55-server & databases/percona55-client now....

Somehow expected there would be more than 12 ports wanting either client or server... probably missing the occurrences in multiline RUN_DEPENDS or some other way to specify the depend. Since pkg_info says there are 103 ports that depend on the client, and two ports that depend on the server (neither being www/owncloud or mail/roundcube, which are the ports that I'm running on 'zen' using the mysql server. On cbox, there are 73 ports that depend on it, some are obviously true...like net-mgmt/cacti and net-mgmt/cacti-spine, but nothing is depending on the server...though it is obviously being used by cacti. I left dbox with the default databases/mysql55-client...there are 71 ports depending on it.

Meanwhile I have postgres server running on zen, which was a depend of something else that I had since removed....but I haven't stopped or removed postgres yet....

In fact after dealing with ruby, flac and perl....the only ports left to update are:

dialog4ports-0.1.3 < needs updating (index has 0.1.5_1)
freeglut-2.8.0 < needs updating (index has 2.8.1)
portmaster-3.16 < needs updating (index has 3.17)

Might be time to see what else I can get working under wine.

Wonder when I had last updated ports in the /compat/i386 on my system at work? And, do I want to tackle that from home, on a Sunday instead of other important tasks/projects....

But first...lets see what port got updated since yesterday....

ports-mgmt/portupgrade

Odd, I seemed to have missed updating to lang/ruby19 from lang/ruby18 on cbox and dbox....

06/29/13

  08:32:00 pm, by The Dreamer   , 2306 words  
Categories: ReplayTV, Time Warner/Cox Cable, TiVo HD / Premiere / Elite, Cox HSI

Cox in Manhattan, KS has gone SDV

Sometime around mid-May, I think, I had gotten a plain business size envelope with just the Cox logo in the top left corner and my address on it. No other indication as to what the content was, so basically resembling similar letters pitching new services or such. Though those will probably start appearing again when the students return....

Not sure why, I opened it for some reason...and eventually found it to be a letter informing me that Cox is about to go SDV and that I would need to obtain some 'free' tuning adapters (where the only 'free' part is no new monthly charge to my cable card service -- where I only needed 4 tuning adapters to go with the 6 cablecards I'm renting.....the two older TiVo HDs having 2 S-Cards each...had thought about switching both to single M-Cards, but given my experience with previous 'self-install M-Card, turning into a full-service install charge because the M-Card was defective....so a tech had to come out to make that determination and give me a non-defective one.)

The letter said that I could visit a Cox Home Solutions store or call to have them delivered. After some thought, I decide that I will call and have them deliver 4 tuning adapters versus no idea how big or when I would make it out that way. Memorial day weekend was out, as was the first weekend of June for sure. And, the switch over was to be like June 25th.

So, I call and delivery is offered to me, which I confirm that is what I want. Where the agent checks to make sure they do shipments to Manhattan, but then later on its apparently not possible to deliver. But, after some further investigation by the agent, it seems like it might be possible and eventually I get a call back saying that they have been shipped and I should get them around Friday (June 7th).

From past experience, items are shipped FedEx Home Delivery, so I had expected to find them sitting there waiting for me when I got home from work...but there wasn't anything. Perhaps Saturday instead....while I didn't intend to wait all the way to when they did show up.... cacti consumed more time than I had intended that Saturday...

The delivery was kind of annoying...in that the delivery person dropped the boxes loudly on my door step with a big whomp...and then it was like he tried to kick in my door before running off.... mainly because it set off my burglar alarm.... &#58;&#35;&#35;

Anyways what I found were 4 huge boxes....big enough to contain DVRs or like components. Perhaps its the only size package they have for deliveries. But, after a quick unpack....the still packaged 4 tuning adapters would fit into one of the boxes, minus the ineffective foam insert...because the insert was sized to protect something bigger than the small boxes that had come instead.

At first it looked like I wasn't getting the complete self-install kit...but later I found that the self-install box contained one coax, splitter and filter...while the tuning adapter box contained the other coax along with usb cable and power wart.

Though I had decided before getting this that I was going to go with a splitter closer to wall and run new longer run to tuning adapter by each TiVo.... And, it was my intent for June 8th to be a Target run. It didn't pan out, and its not something that is sold by the union computer store or the union bookstore...so I ended up ordering cables from Amazon.com.

And, then it was a matter of figuring out when I would get around to setting them up. A number of times where I could've done it went by, but it didn't occur to me that I should do it then. Since the operation would be disruptive to my cable hookup, needed to avoid primetime. And, needed to be sufficiently awake and steady too.

And, then I notice there's a $10 self-install charge on my Cox bill.

Full story »

  05:53:00 pm, by The Dreamer   , 1493 words  
Categories: Hardware, BOINC, Operating Systems

Another Radioactive@Home sensor in Manhattan, KS

So, there's this BOINC project out of Poland called Radioactive@Home, where you have a radiation detector hooked up to a computer taking samples, etc. Its my second BOINC project with a hardware sensor. Though I had signed up for this one first...back on June 16, 2011. QuakeCatcherNetwork had come later, but getting a sensor was quick (though there were delays in getting it working, they had switched to a new sensor where they didn't have Linux drivers yet...etc., etc.) But, doing Radioactive@Home took longer as sensors are built in batches, there had been early batches that I missed and I wasn't all that sure at first if I really wanted do go to the hassle of getting one.

But, then another user announced that he would do a group purchase of 50 or so, which it should cut shipping costs quite a bit by having a cheaper large shipment from Poland, plus domestic delivery for the last leg. The way delivery costs go, you can get up to 3 for the delivery charge...though most people only want one....at least initially.

Basically I ordered my first detector around August 2011, and finally received it in March 2012. And, it just runs...though occasionally I'll look to see if anything interesting is recorded (like the interesting trace for around the end of the world....)

Meanwhile, on June 26, 2012 there was an announcement of a new detector...a pretty looking one. My first sensor was a prototype type case with rough cutouts, etc. Not really bad looking, but still plain and crude looking. While the announced sensor looked neat, the kind of thing that I might considering putting on my desk at work....

So, there was basically an announcement that there wasn't going to be another bulk US purchase...so after some thought, I decided this new detector was just too pretty to pass up. So, I ordered one mid to late July, 2012. Got confirmation on July 23rd, 2012. 27 Euros for the detector plus 10 Euros for up to 3 detectors, more than 3 pay for the detector now, get bill for actually shipping cost later. Plus if I use PayPal to specify that I'll pay the transaction fees....

In the previous order, it had been requested that we have PayPal funds to pay for the transaction....or use a check. I had tried to keep a float of cash in my PayPal account....but when it finally came time to pay, there wasn't quite enough to do that, so I opted to just mail a check. For this second order, I went with PayPal and had PayPal add the transaction charges to my total.

First detector cost me $46.25 by personal check. Second detector cost me $47.36 (and conversion and including the transfer charge).... I sent the PayPal money on August 21, 2012.

And, then it was wait and wait and wait. I would check the boards now and then for updates...but it was mostly other people wondering the same thing.

Eventually, I stopped checking in...and kind of forgot all about the sensor. Though I did visit the site briefly, but didn't linger or read the detector threads...which I went to check what platforms the project supported. Because when I had originally ordered, I was down to a Solaris 10/x64 workstation, a Windows box, a first gen MacBookPro (32-bit Core Duo). and a dead Linux machine. Eventually, I got a computer to replace the dead Linux box...but I went with FreeBSD instead, and it eventually displaced the Solaris workstation. In February, 2013 while I was working late on my FreeBSD system, I saw the Windows box update itself and reboot, and then it failed to boot. It had killed itself....pretty much the same way my home Windows box had killed itself in an auto-update in February, 2012. I left it off, not sure what I would do with it....I thought about OmniOS or SmartOS...though it was a first gen i7, so no EPT for KVM. Eventually, I decided to install Ubuntu 12.04LTS on it....where its mainly backup for when my FreeBSD system crashes.... its one thing that new Seagate drives only have 1 year warranties...its another thing that they seem to have trouble lasting that long.....

And, then an iMac 27" appeared on my desk....back when it seemed bleak on getting FreeBSD working as my main workstation....I was talked into getting one. But, FreeBSD remains my main workstation....while there are somethings that the iMac is the only computer I have where things work (like being able to participate in WebEx, Lync, Google Hangouts or Xoom for web conferencing....plus it finally solves having mail staying open while I switch to the appropriate desktop to do whatever....I'm up to 17 now....where there are typically 4 to 12 windows...either of uniform size, or variable size, and some desktops the windows overlap, though that desktop is mainly for tailing logs.... Where I'm up to 2 full desktops and 2 half desktops for that.... Anyways, I had made a quick visit...because I wondered if Mac OS X was a supported platform (it wasn't) or if anybody was using FreeBSD for this project....didn't get any search hits. And, it seemed unlikely that the hardware part would work through the Linux emulation on FreeBSD (especially the Fedora 10, and I'm not sure what the process for converting to the CentOS 6 is, that wouldn't break all the things I'm using Linux emulation for....though it is mostly other BOINC projects.) Though doing the search now, I see that a couple days ago the question got raised....with not much luck on having it find the detector ... but ending with a link to a FreeBSD version of the application.... Though since I have a Linux system at my desk (where is primary purpose is to run VBoxHeadless containing Windows 7, for those occasions where I need to use vSphere Center...and passing the time doing BOINC)...I'll just go with running new detector should it ever arrive...on that.

Full story »

06/24/13

  08:57:00 am, by The Dreamer   , 838 words  
Categories: Home Theatre, Software, Momitsu V880N, FreeBSD, CFEngine

This just in, cfengine developers don't test or use cfengine!?

So, this morning I was was wonder why my nagios was still warning about something that it shouldn't be. I was positive I had changed the warning threshold above where it was. I do an 'svn status' on my work dir, nothing uncommitted. I do an 'svn up' on the cfengine server....no updates, I drill down to the file and its correct (perhaps I need an alias on this side as well...though I usually only use 'cdn' for where my svn work dir is or on the nagios server....though its because at work....where this alias is used in association with nagios as well (where work nagios is not yet managed by cfengine, but was considering it for the new nagios server that I'm trying to set up between fires and stuff at work....except the fact that we're still running cfengine2 is really starting to become a problem......though I wonder if cfengine2 could do it, if it weren't hampered by how former admin had implemented things....The work cfengine made a mess with using it to setup a new system because of weird cross interactions between 'promises' and that the promise wasn't written in the same sequence it was running, things that probably aren't a problem when cfengine was original deployed to promise that nothing ever changes....)

Anyways....I finally hunt through the -v output... which is now not much different than debug noise, and nothing like what verbose used to be in 3.4.4.....no more search for 'E nagios' to find where the start of "BUNDLE nagios" is in the out, and then finding the specific file promise..... what a mess. Its like they don't want you to know what's going wrong....

Turns out I missed some more uses of 'recurse' from cfegine_stdlib.cf, where xdev=true is busted.

It was one of three bugs that I had logged for cfengine 3....#2983. Which was almost immediately flagged as a duplicate of #2965 (3.5.0rc fails to recursively copy files with strange message)...and this morning at 5:03am, my bug was closed as that it indeed seems to be fixed for 3.5.1 (soon...).

Wonder what the definition of soon is....had a previous problem where cfengine was complaining about bad regex....when the default for insert_lines: is that they are 'literal' strings. Which was making it hard to use cfengine 3.4.x to make edits to my crontab files. After putting up with it for a couple of months, I finally visit the bug tracker and find that its already been reported and fixed for next version. But, months and months go by and no new version appears. Though it does seem to be fixed in 3.5.0.

Anyways reading #2965 was interesting.... aside from where the dev? spots another bug in the same code and has that pulled as part of the bug. Also that it was reported against RC, and made it into release. Though I had reported a bug in against an ubuntu 12.04 beta release....and it persisted into the release version, where they debated fixing it because apparently LTS means don't update anything after its release...(though I thought they had said things like firefox would stay current instead of staying fixed at the version at time of release now...) Plus it seemed I had to keep reminding them that my bug was reported before release, so that should be reason enough to release the fix. I'm pretty sure they did, but I hardly use that ubuntu desktop anymore (or any ubuntu desktop....though I did fire up my laptop yesterday, but its because there was a new VirtualBox and I hadn't updated the XP VM on there in quite some time....though I've been thinking of whether a FreeBSD laptop is feasible.)

Someone asks that they have a unit test for this bug. Where the response is a unit test would need a running server, which they don't have (yet)...how long has cfengine been around for them to not be using it? Sure wouldn't want to be somebody who's paying for this.

So does that mean nothing is being tested, and that nobody involved in development use cfengine? Because this was the kind of bug that pretty much anybody that uses cfengine3 would run into. Considering I only have the 3 systems (zen - policyserver, cbox, dbox) at the moment....

Perhaps I'm jaded by having worked for an Enterprise software company and how we did full builds every week, and with full runs of automated and manual QA testing. And, having to create unit tests for less than trivial bugs as part of fix/review before closure process. Though what I'm hearing about Chef...its worse....

Still haven't decided what I'm going to do with my Linux systems....migrating the files from Orac if I were to turn it into FreeBSD is the stumbling block, plus I would lose certain services...some of which might not really be an issue, since its probably time I make the leap to blu-ray. And, either I get another Roku or figure out how to incorporate the smart side of my TV into my life (probably time to finally upgrade my receiver....purchased October 27th, 1999)....

06/21/13

  06:52:00 pm, by The Dreamer   , 522 words  
Categories: General, Healthcare

A shift in my vision

After I got my new glasses after my eye exam back on February 8th, things were great...but gradually they started getting back again, which was some what distressing. Especially since I had also gotten that call about the results of the Retinal Thickness Analyzer (?) showing some thinning and that I needed to plan a follow up test in about 6 months. At first the vision would start getting bad as I was heading home from work, and it subtle...though it would get pretty bad about mid evening.

But, then it was starting to get pretty bad around mid afternoon at work, so I decided I needed to get my eyes rechecked. In the inbetween time of making the appointment and the appointment today, I now feel that my vision starts getting bad before I leave for work, sometimes wonder if I should even bother (though I end up going in anyways and struggle until quitting time...)

So here's the next entry for my prescription progression...

Nov 2003: OD -5.00 -1.00 075       OS -8.00
Apr 2005: OD -4.25 -1.00 080       OS -7.00 -0.75 045
May 2006: OD -4.50 -1.00 055       OS -6.50 -1.00 035
Apr 2008: OD -5.25 -1.00 060       OS -7.25 -1.00 025
Aug 2009: OD -5.00 -0.50 050       OS -7.00 -1.25 030
Oct 2010: OD -5.50 -0.75 050 +1.25 OS -7.00 -1.00 040 +1.25
Nov 2011: OD -5.50 -0.75 050 +1.50 OS -7.00 -1.00 040 +1.50
Feb 2013: OD -4.50 -0.75 040 +1.50 OS -6.75 -0.75 010 +1.50
Jun 2013: OD -5.25 -1.00 060 +1.75 OS -7.00 -0.75 040 +1.75

Somehow my eyes are close to back to where they were in Nov 2011...

Nov 2011: OD -5.50 -0.75 050 +1.50 OS -7.00 -1.00 040 +1.50
Jun 2013: OD -5.25 -1.00 060 +1.75 OS -7.00 -0.75 040 +1.75

There's a slight change in the right eye, and even less of a change in the left eye....reading power another step up....

Seems a really strange jump and not really sure how to explain it. In what feels like an unusually rarity in my life, there hadn't been any medication changes from when I started having problems last fall to no, and no new diagnosis....though the latest problem that I've yet to receive a diagnosis for started mid-October....and it in November that my vision started getting bad (though with LISA and my mom's 70th birthday coming up, and thinking that I wanted to wait until I had FSA money...I had put off getting the eye exam. Plus it wasn't as bad as it has been lately...

As to why the shift...they had done a cornea surface mapping as they have done in all previous exams....and in comparison...my cornea surface was much more pointed in February while my current surface is similar to how it was back in November 2011. Not really sure how that came to be though. But, Optometrist said he would order up new lenses and that there would be no charge. Hopefully, they'll be able to get new lenses already pre cut for my glasses, as these are now my only pair....not having kept any of my old pairs this time around given the big change. (and not yet having gotten around to getting a second pair, because around when I was starting to think that I should get them...was when my vision started getting bad....)

Not sure where I'm at on getting a second pair now....will probably shoot for getting them before the upcoming NN Conference for sure.

Hopefully its not some kind of periodic variation, possibly similar to the variation in my sleep....which appears to still exist even with my Narcolepsy being pretty well controlled with medication now.

  10:18:00 am, by The Dreamer   , 181 words  
Categories: Software

Upgraded tardisi.com to zenphoto-1.4.4.8

What a pain that upgrade was....it kept change files and directories to really weird permissions. Like setting perms on index.php to 0, or themes directory and the themes/default directory to 311 (d-wx--x--x), and changing the parameter for CHMOD_VALUE to an $conf['CHMOD'] which is undefined...from its previous value of 0755. There were other places it had changed to perms to 0, etc. It was also complaining that it wanted log files to be 0600, but they are set to 0600...except for debug.log which it had changed to 0.

Eventually, I cleaned up the mess and as soon as I got the setup to show the go button, I clicked....ignoring the other warnings...such as mysql with hosting provider meets minimum requirements but not preferred. or that it thinks files already set to strict perms is wrong and that I should set them to strict (which was part of why things were being changed to no perms at all....)

Things seem to be working again.

I think I should get back to hunting down photos and filling in the empty albums, and see if other photo sets should go up....

  08:48:00 am, by The Dreamer   , 150 words  
Categories: FreeBSD

amd[###]: No map entry for <old_share>

For almost a year, every 3 seconds in /var/log/daemon.log there are two lines:

amd[####]: No map entry for <old_share1>
amd[####]: No map entry for <old_share2>

Every now and then, I would spend time hunting around my filesystem trying to figure out what part of amd might still be holding onto the old information...its not in any of the config files and not in any of its directories, its not making symlinks for it anywhere.

I was about finally submit and try asking on a mailing list....when while I was compiling how I did things into the message, I expanded my search to make really sure that there weren't any lingering references....when I found that I had two symlinks in my homedir pointing to the old shares.

Once I removed those symlinks....the messages stopped.

I guess the gam_server process that is watching my home directory was polling those symlinks every 3 seconds....

&#58;&#111;&#111;&#112;&#115;&#58;

06/15/13

  04:04:00 pm, by The Dreamer   , 1123 words  
Categories: Software, FreeBSD, CFEngine

cbox/dbox cfengine update also full of fail

First I had saved libpromises.so.1, so that I could invoke cf-agent from /var/cfengine/bin to pull in the new cfengine-3.5.0 binaries and pull up the new inputs from my policy server.

Except I forgot to commit the 'bundle agent foo' kluge, and I had done an 'svn revert ...' to undo all the fiddling I had been doing on the policy server.

But, after I make the change, cbox/dbox refuse to copy up the new 'do-crontab.cf' file. I try running things verbose, 'cf-agent -v > out', but there's no out file??? &#58;&#63;&#63;&#58; Did I slip? Am I losing my mind?

Guess I should do it in another directory, because the update does purge...so its removing my 'out' file. &#58;&#111;&#111;&#112;&#115;&#58;

So, its saying this:

2013-06-15T14:40:55-0500  verbose: Entering directory '/var/cfengine/inputs'
2013-06-15T14:40:55-0500  verbose: Device change from 1242801830 to 843968349
2013-06-15T14:40:55-0500  verbose: Skipping '/var/cfengine/inputs/do-mysql.cf' on different device
2013-06-15T14:40:55-0500  verbose: Device change from 1242801830 to 843968349
2013-06-15T14:40:55-0500  verbose: Skipping '/var/cfengine/inputs/do-ddclient.cf' on different device

What are those device numbers....is it saying that because the file is remote that it won't copy it? After some looking around....I see that 'u_recurse("inf")' is:

body depth_search u_recurse(d) {
        depth   => "$(d)";
        xdev    => "true";
}

Which seems perfectly reasonable that I want it to recurse the destination directory and not cross devices. But apparently, it now means don't recurse into devices that different than the source directory's device???

I look around to see if I had done the 'xdev' line or if that was from where I based my initial setup from....and find that its what was giving over on Unix Heaven - http://www.unix-heaven.org/node/53#cf3-update

That change in behavior totally doesn't make any sense? But, enough hair pulling.... let's get things running again on cbox & dbox.

Hmmm, now its spewing warning messages:

2013-06-15T14:47:31-0500  warning: Setting replace-occurrences policy to 'first' is not convergent
2013-06-15T14:47:32-0500  warning: Setting replace-occurrences policy to 'first' is not convergent
2013-06-15T14:47:32-0500  warning: Setting replace-occurrences policy to 'first' is not convergent
2013-06-15T14:47:33-0500  warning: Setting replace-occurrences policy to 'first' is not convergent
2013-06-15T14:47:33-0500  warning: Setting replace-occurrences policy to 'first' is not convergent
2013-06-15T14:47:35-0500  warning: Setting replace-occurrences policy to 'first' is not convergent

I didn't before...and there's only one occurrence, so it is convergent. And, my use of 'replace_with => With("...")' for 'replace_patterns:' was lifted from -- https://cfengine.com/archive/manuals/cf3-solutions. So, in With(), I should change 'occurrences' from "first" to "all"....why have this attribute if its purpose is just to annoy now?

Its already annoying enough that -q is gone....but -K remains? If I'm running a cf-agent manual to speed up picking up a change....I'd like having to wait splaytime, but I don't want to ignore locks...because if I'm close to when cf-execd wants to run....it results in a big mess, especially for promises using single copy nirvana.... cf-agent doesn't distinguish on why the specific copy failed when there's a collision in this case, so it causes a less specific version to get copied....so that it has to fix it in the next run.

This causes problems for things like DNS and NTP. The specific is cbox is an NTP server that polls external servers in the freebsd.pool.ntp.org....and the generic NTP config is to use cbox and dbox as NTP servers....which has resulted in wild client oscillations, though both my dsl and cable connections have been rather unstable lately...though dsl will just have periodic drops, cable goes out until I reboot my cable modem....and the fact that using two NTP servers is probably the worst combination. But, I don't think I could get away with 4+ NTP servers on my home network. DNS....well, specific is cbox is primary authoritative, dbox is secondary authoritative, and generic is recursive caching resolver....similar to how I did DNS servers under cfengine at work.....

4+ justification - http://www.ntp.org/ntpfaq/NTP-s-algo-real.htm#Q-NTP-ALGO

NTP likes to estimate the errors of all clocks. Therefore all NTP servers return the time together with an estimate of the current error. When using multiple time servers, NTP also wants these servers to agree on some time, meaning there must be one error interval where the correct time must be.

Number of servers:

  1. Always trusted, even when its totally wrong
  2. If they differ, how do you break the tie?
  3. A server failing results in above.

So, 4+ servers for reliable accuracy.

Whether the total should be odd or some other sequence, is another debate.

http://www.ntp.org/ntpfaq/NTP-s-config-adv.htm#Q-SERVER-NUMBER

But three is a good place to start, and you can progress to three-groups-of-three if you feel the need.

three-groups-of-four? four-groups-of-four?

Hmmm, I don't know why I had never done 'cf-agent -v > out' in /var/cfengine/inputs on cbox/dbox before....but often do it on my policy host. I guess I don't normally need to be in /var/cfengine/inputs on cbox/dbox when I'm troubleshooting cfengine.

Hmmm, so the verbose run, I only see two places with warnings??? That seems odd.

cf3>     .........................................................
cf3>     Promise's handle:
cf3>     Promise made by: "name="tor""
cf3>     .........................................................
cf3>
cf3>  -> Looking at pattern name="tor"
cf3> WARNING! Setting replace-occurrences policy to "first" is not convergent
cf3>  -> Verifying replacement of "name="tor"" with "name="tor2"" (2)
cf3>  -> Replace first occurrence only (warning, this is not a convergent policy)
cf3>  -> Replaced pattern "name="tor"" in /usr/local/etc/rc.d/tor2
cf3>  -> << (2)"name="tor2""
cf3>  -> >> (2)"name="tor2""
cf3>  -> Replace first occurrence only (warning, this is not a convergent policy)

Wait....it makes even less sense.... this edit promise is about creating /usr/local/etc/rc.d/tor2 from /usr/local/etc/rc.d/tor. Since the starting file is fixed, the first occurrence is inherently fixed...so replacing it is convergent. So, its just there to annoy.

And, wait....its also having issues using recurse() from cfengine_stdlib.cf....well, it has xdev => "true"; set as well. And, that body is unchanged from the previous version.

Oh, I suppose I could remove splaytime.... its not like that I have a huge number of systems....

Wait... the tor2 edit_lines also has an if_elapsed("60"), so warnings are from other edits that were probably skipped. Guess, I'll have to risk a -K to find all the occurrences of the warning.

Though its strange that I'm only seeing the messages on dbox, even though the promise exists for both cbox & dbox....

OTOH, on cbox it did catch an error.... where it was promise that "/usr/local/www/apache22/data/tivo/.", etc. was accessible to the webserver Except I had typed "/usr/local/www/apache2/data/tivo/.", etc. There was no error/warning about skipping this promise before. Though in 3.5.0, it gave an error that it couldn't chdir to this directory...which delayed me into noticing the actual reason for the error.

And, once again...another wasted Saturday.... I didn't really need to go to the mall and spend money I don't have, I suppose. &#58;&#104;&#109;&#109;&#58;

Now before I can go back to just doing 'portmaster -a' to keep current....I still need to decide the fate for the databases/mysql55-server update.... I suppose while cfengine was down, I could've done the upgrade if the decisions was to stay put. But, still wondering if I want to try percona, or perhaps better yet, maria! I wonder what my friend Maria is up to these days....

  01:15:00 pm, by The Dreamer   , 694 words  
Categories: Software, FreeBSD, CFEngine

Meanwhile upgrading cfengine-3.4.4 to cfengine-3.5.0 not going well.

Upgrading the port was no problem....but it broke my cfengine. Why? The port puts the cfengine binaries in /usr/local/sbin, while the cfengine practice is that it has a private copy in /var/cfengine/bin. Which would be fine if the binaries didn't have shared library dependencies. Which they do, specifically libpromises.so.1 which is gone in cfengine-3.5.0...there's a libpromises.so.3.

Though before I discovered this problem, I first wanted to make some tweaks to update.cf so that I would have some indication that it had copied up new binaries from /usr/local/sbin to /var/cfengine/bin, since I noticed that files there newer than expected. Though I probably just rebuilt the same version port because a dependency had updated and /usr/ports/UPDATING indicated that I need to do that.

This probably is why at work, the person that setup our cfengine 2 went to extreme effort to create static cfengine executables...ignoring that such things are officially not supported on Solaris. Though we seemed to get away with running those executables, built on a Sol10u3 sun4u system...on systems more current up to Sol10u11, and a few Sol11 systems and systems that are sun4v architecture.

In a past life...we had run into a statically built executable (the installer) not working on our first UltraSPARC-III system (Sun v280r)...trying to recall what our build machine was back then.... my recollection says we only had the SPARCserver 20 and SPARCstation 10, before that. Though as I recall, we had to wait for a patch from Sun as well as rebuild the executable shared on the SPARCserver 20...to have it work. It wasn't long after that though that we retired support for sun4m, changing minimum requirements. Wonder if the application has become 64-bit yet? But, for ABI backwards compatibility claim to work, the executable needs to be built shared...so that it'll find the libraries provided on newer systems to allow older executables to still work.....

portmaster probably didn't know that it should save /usr/local/libexec/cfengine/libpromises.so.1, though would the old executables know how to find the library when its moved aside? (I do have SAVE_SHARED=wopt uncommented in my portmaster.rc file).

Occurs to me that I could just restore the file from backup, it would allow me to run

failsafe.cf

and get me to where everything should work again.

Though before I did that, I had invoked cf-promises (the one in my path -- /usr/local/sbin), and it complains about library.cf. Guess it doesn't like the old cfengine_stdlib.cf, the new one isn't where the old one was....it was here instead --> /usr/local/share/cfengine/CoreBase/libraries/cfengine_stdlib.cf I do a quick look at what's in it....mainly to make sure that bundles/bodies that I use are still there...and notice some interesting new ones....such as a package_method of freebsd_portmaster, someday I should look at cfengine3 to do port/package promising....

But first get cfengine working on policyhost, hopefully the other servers (at 3.4.4) are still working.....guess not, 3.4.4 doesn't like the 3.5.0 cfengine_stdlib.cf file. But, cf-promises is also not happy with some of my other promises....

Guess I'll update those while I get policyhost working again.

.
.
.

Or perhaps I need to revert....

root@zen:/var/cfengine/inputs 317# cf-agent
2013-06-15T13:22:53-0500    error: Bundle 'crontab' listed in the bundlesequence is not a defined bundle
2013-06-15T13:22:53-0500    error: Fatal CFEngine error: Errors in promise bundles
1.755u 0.113s 0:01.94 95.8%     172+2501k 133+12io 1pf+0w
root@zen:/var/cfengine/inputs 318# 
# cf-agent -v
...
2013-06-15T14:00:57-0500  verbose: Parsing file '/var/cfengine/inputs/do-crontab.cf'
...

Its there, why's it not working.... 'cf-agent -d' doesn't work, but it will only do failsafe....

Full story »

  09:12:00 am, by The Dreamer   , 390 words  
Categories: Software, FreeBSD

Perl update continued

For some reason I had cd'd into /usr/local/lib/perl5 on dbox and noticed that 5.16.2 was still present...well 5.12.4 was still on zen after I the upgrade to 5.14.2... and it just had whatis files. But, I went and looked inside, and found more than just whatis files.

Using 'pkg_info -W', I found that I had other ports that had installed perl modules that didn't start with 'p5-' or depend on 'libperl.so'.

So, off to rebuild those ports.

On dbox/cbox it was just databases/rrdtool, print/pdflib...plus some stray files left by already updated ports or removed ports. But, on zen there was a much bigger list of ports:

security/clusterssh
graphics/ImageMagick
japanese/p5-Jcode (which was missed, because the package name is ja-p5-Jcode-)
devel/perltidy
mail/razor-agents
security/clamtk
print/foomatic-db-engine
graphics/gscan2pdf
x11-clocks/intclock
databases/rrdtool
print/pdflib

Hmmm, probably need to update my i386 space, which is going to be wrong now...because uses the name make.conf of 'global', and I haven't updated it in a long time.... Not since May 4th. emulators/wine-devel has been updated since then, so I guess I'll have to tackle it sooner than later.... Especially, since I'm thinking of making another attempt to see if I can get other apps running in wine versus VirtualBox....

Full story »

1

Now instead of subjecting some poor random forum to a long rambling thought, I will try to consolidate those things into this blog where they can be more easily ignored profess to be collected thoughts from my mind.

Latest Poopli Updaters -- http://lkc.me/poop

bloglovin

There are 20 years 1 month 28 days 15 hours 32 minutes and 23 seconds until the end of time.
And, it has been 4 years 10 months 29 days 22 hours 30 minutes and 33 seconds since The Doctor saved us all from the end of the World!

Search

June 2013
Mon Tue Wed Thu Fri Sat Sun
 << < Current> >>
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
Google

Linkblog

  XML Feeds

Who's Online?

  • Guest Users: 0
This seal is issued to lawrencechen.net by StopTheHacker Inc.
multiple blogs

hosted by
Green Web Hosting! This site hosted by DreamHost.

monitored by
Monitored by eXternalTest
SiteUptime Web Site Monitoring Service
website uptime