Posts made by Foglalt

@Tom-Elliott omg, i missed to notice it? bah, shame on me. maxima mea culpa! i go to the corner and undergo my punishment

Sometimes ppl fail, but recovery can be done on almost all situation somehow. It needs many factor and one of these factors is the properly logged events. I dont want fog to become a logging system, but sometimes mysteries can be less disturbing if we can see more logs.

Some of the staff member deleted a few images to clean space and maybe mistakenly deleted wrong image. The extra and most serious problem is that it took too long time to realise it and there came the backup ( well, we have a lot of images and to keep the usage at bay, we cant take too many instances of an image, so backup is the worst for such situation: images are only replicated to a different location after a time passed).

Normally a new image when taken has unique name, so this backup strategy is ok with it (nothing overwritten). Deletion is extra rare. Image retake to same name is not permitted and if somehow done, from backup it can be rewritten with the original file. This is the weakest point: if time goes and no one notices, it can be overwritten in backups, too.

On limited space no “proper” method, only “best of possible” is available. But to recover disaster (or at lest shock of wth happened) I would want to see logs of events like:

• host/image creation (who when what)
• host/image deletion (who when what)
• host/image overwrite (who when what)

There is imaging log which i can see, but the above mentioned ones are sometimes even more important.

Can it be somehow invented?

@Junkhacker i will inspect the postdownload script section usage (well, documentation search in some topics is hard for fog system, for me at least). but atm we have a lot more disturbing problem, mystery that keeps me away from sleeping. topic is “missing image”

@Junkhacker I have such ideas, partially went through it to see if the concept is working or not. thx for comment maybe somehow i will figure out a solution. With our situation the main issue is to have a document after it with deletion (like vertificate), and make it as easy as possible to have untrained staff to use. “premade image” is a “light” solution. fog’s disk delete feature is a “semi-ok” solution. active kill disk and such softver is the ideal (only it is not properly automated in a mass way).

If i have enough free time (non-existent in my dictionary) i will find the proper way.

@Wayne-Workman Actually no full automation is needed. I just want to hand out job to ppl who are not needed to param a util properly with 10 or so options every time. Many occasion to mistype, etc. pxe menu is used, but the selected software needs tuning. I was just wondering about a documentable method utilized maybe from fog’s pxe environment. And maybe do some documentation with the process automated.

thx for cooperative thoughts anyway

@george1421 I absolutely have right, but you misunderstood me a bit. Let me clear things up a bit. IT wants to “kill some old pcs” and mass delete their contents. Now this procedure is slow, but goes like this: pc comes from basement storage rooms, identified to delete. Its cables are connected, boot set to usb, from which a scripted killer software runs to kill disks. Notes taken (old school, paper etc), then documents created later to achive about the process.

We are looking for a better method, something like this: pcs are gathered, identified and so on, all connected to network. Then task is given, maybe over pxe environment, and crew (DONT have to be IT staff, ppl who can connect cables, watch screen to see if it is done) wait, disconnect pcs after work, then go for another pack of pcs.

We dont want it like this: registered pcs in fog as hosts and at a certain time select for deletion and commencing it. (actually we dont have permanent hosts in fog database, only lab computers. others are just regged only for imaging purposes, then removed).

Accidental “boss pc kill” can only be done if registered that mac, set task, reboot and NOT NOTICING it is not started as task, wondering what happened, then forget it without deletion of false task. and later boss pc turn on, disk killed. No, it is absolutely not happening this way

Still you see risk in this theory or is it cleared? (btw I appreciate that you wanted to point a possible problem source!)

Hi!

We are using fog system to create and deploy images and now we are planning a mass killing of pcs (all harddisk). Normally we used softwares like Active Killdisk 8well, actually we test it how to and what to. We bought a licence to test. It can create bootable iso to have a semi automatic method. With it I can use pxe environment to boot that from network, it does scripted killing (i hope we can make a properly macroed image to zero interaction; anyone use it maybe? ).

I dont want to use pxe menu manually to start it. Can I somehow tell for to boot something which is in its advance pxe boot menu? Lets say I make an iso, put it in the pxe boot menu as a new entry, test it, then use it like a host task from Fog gui. If so, i can give a pile of pcs to not-trained workmen to systematically put pcs together, boot them, wait till it is killed, switch off, get new ones, etc. I would like to use as less interaction as possible.

With fog i can kill a disk, but i cant have it “proved” (or can i?). With active kill or similar softwares there can be created a certificate (date, serials, etc in a pdf). If i make the mentioned iso to create a file, somehow gather information with fog maybe (with post image scripts or so).

Is it to absurd to do it? have better ideas? I am open to anything. Onlything is needed that i can make those certificates as much automated as possible with as less knowledge of the operating crew as possible.

(As a random idea that bugs me, can i use a premade “image” to overwrite hosts.normally hosts have 1 disk only, but what if they have more…)

@Junkhacker thx for comment

Yes, the accuracy thing is same as we thought. The peeks anyway maybe real ones. The maxed out part is real, I mean no other activity (no replicating, etc). It was dont in isolated time, multiple occasions. That is why it looked strange. Next week we try to get other tests with different, maybe dumber switch to see what goes on. I will report in what we found. If no other changes on fog side of multicast, only inbetween hw can be the reason I think.

We are thinking about the “newly” installed managed switch is the source of the problem (we have such ages before in one of our sites). This one was setup in hope it can solve “allways there is a need to a new slot on the wall” problem. It is set for like this: some pc, some phone, 1 nas (all 100 mbps but nas has giga). and some of its ports are set to gigabit for purpose of fog usage. We suppose that it replicate traffic to wrong directions maybe which it should not, generating network issue. That will be killed somehow I hope soon (well, next week is the closest).

But those peeks are strange and disturbing. Why are there those at all? And why it is maxed out to 100mbps, if those link for requestors are gigabit? Throttle control on fog? (both end has gigabit slot, cpus are not overloaded, network is mostly still at this time).

Anyway, there was rumours about torrent imaging? Will there be once?

Sorry, i wasnot aware that I didnt press post on my reaction to this topic (pc left on when i went home and family killed memory fagments :D)

Yes, missing host issue was fixed. After that i could register that host again! THX for cooperation and kind help!

Anyway can i “mark it solved” somehow?

Hi!

We are experiencing strange network issues which was short and random “bottleneck” effects. At first we thought it is an outside source of problem (too short to be able to detect) but slowly it led us to fog activity as the source.

As we cut down to the bottom of it finally we could reproduce it this way: 2 machines imaging in multicast setup. Server and wall socket are on gigabit link, but on the “client side” a managed switch, patchcables to clients from that. When imaging starts serverside bandwidth is going to fill up to 100mbps (strangely not to 1gbps) and oscillating around a “max of 100mbps”. Few sec from start, network lags occure. Even ip-phones drop line or go choppy, which is unique.

This effect comes and goes during the imaging, but mostly 2-3 occasions. We will go deeper investigations (get different switches for exaple, etc), but i decided to show the pattern of it on the serverside “gui of bandwidth utilization”.

Images shows that from maxed out 100mbps sometimes there are peeks to even over 1,5gbps. Those should be false values maybe, but signs where and what happens. Those times phones struggles to be alive, streams or real time connections drop off, etc.

Why would multicast do such? For some other reason we stepped from 1.2 (that time we dont have this for sure) to 1.3, then 1.3.4 of fog versions. Unicasts dont do this effect.

Is there anything that would give acceptable reasons for those peeks? And does fog have throttle control somehow?

Just for your intention: that strange user and password that was found in our database is even more strange. Those are actually fog users and passwords saved with host data. Fog client was never used in our environment and all (which would require ad passwords maybe).

Even more, user for ad cannot be saved from host or image gui part. None of my colleagues use any of db manipulating software (actually never ever used for any purpose ). This means that those fields are mysteriously got those info!

This database was created with version 1.2.0, previous db was not used at all only as reference for the new database building (hosts are mass imported from newly created lists coming from other source).

So, if our database is corrupted somehow, we dont have no clue how it was done.

@Wayne-Workman thx for the work. About primary. Can it be checked somehow by code that a host with no primary mac cannot be save? I think i have a guess how it came.

That host field, the mac is normally updated when new cloning starts (we regularly do few hosts in pack as company now upgrades all pc “one by one”. I will check, but i think it is done as follows: new mac copied from hw library and pasted into for host record. Since now it has primary and other field, maybe colleague did it to other field, not the primary (he only deleted primary, then by mistake copied the new mac to nonpripary field.

If possible, pls dont let it be updated if no primary mac entry exists. Maybe it prevents this. I will ask him how he does, but i fear he wont remember the exact process, as it is dont in “half robot” mode (tons of pcs are repared, etc).

As for the hostADPass field. I forgot to ask it, but inspecting the problem i saw that, too. Is that field is for entering hosts to domain? I was not even aware that they use it. As far as i remember it is dont by scripting on the host, not by fog at all.

@Tom-Elliott No, none did any manual work about that password Those who use fog for imaging, well, dont even care how it works. I will ask them, but i am totally sure they will be surprised even hearing it! (and shocked that password thing)

@george1421 thx for helping me out!

@Tom-Elliott and what does “not official” means?

Btw if zero means it is not pending, how can i get rid of this host? As it has mac, mac is not pending is it a legit host playing hide and seek somehow? Anyway i see one interesting thing. the problematic host’s mac has no primary flag. Can it be the problem? if i modify that from 0 to 1 maybe it comes back to visible? (can it be done with database browser anyway, or is it “rape” of the database integrity?

sorry for ugly output, from phone it is hard seems it has mac associated. what does pending means in this context?

+------+----------+-------------------+--------+-----------+-----------+--------                                                                                                                                                             --------+-----------------+
| hmID | hmHostID | hmMAC             | hmDesc | hmPrimary | hmPending | hmIgnor                                                                                                                                                             eClient | hmIgnoreImaging |
+------+----------+-------------------+--------+-----------+-----------+--------                                                                                                                                                             --------+-----------------+
|  414 |      107 | 9c:5c:8e:93:ac:dc |        | 0         | 0         | 0                                                                                                                                                                           | 0               |
+------+----------+-------------------+--------+-----------+-----------+--------                                                                                                                                                             --------+-----------------+
1 row in set (0.00 sec)

Additional information, until you can suggest me anything how to remove the unwanted “already registered” host i watched the database with a browser to see manually under the hood. host is actually there in hosts table. as i dont know the links of tables i wont delete it but it may help if u need more info on this issue.

so, @Wayne-Workman or @Tom-Elliott (or anyone ofc) what can i do with this “lost” host in the database?

@Wayne-Workman

I did upgrade, but sadly it says that “hostname exists already”. as i am not an sql pro, can you pls tell me what else can be done to find reason fo it?

one more thing. as it first occured the colleague who found this issue uses his own “clone hosts” and when imaging comes, he only renew MAC of the new hw, changes image if needed, but hostname or host “item” is not deleted. it sounded absolutely working, but does it any wrong? (in my logic it doesnt, but pls confirm me)

Case reopened. I still use version 1.3.0 (well, have no time to test all versions in the last few subversions were fast to come out). Last time i could not register new host cos it said mac already used. not it says hostname exists, but i cant find it. (well, the new host lister part in gui is strange for me, but other hosts it list i search for but not this particular one).

Do you maybe have another tip what is it? (i did what you suggested last time, but now it “affected zero rows” now).

We are over some tests at last. As you suggested I did the following:

• used a total dumb “link” between the pc and the network (hub): no change, still no boot properly
• I used undionly.kkpxe and ipxe.pxe one by one. unfortunatelly no change.

Still no change, that means this hardware has somehow problems. Any other suggestion maybe? The problem is that we will surely get such hw, so it would be good to have a workaround somehow. Ah, one more thing. ipxe.pxe file gave a little difference. It said during boot that it boots from SAN device. Does it changes a thing?