My company still uses JDE software, running through old AS/400 console emulators... Yeah, big companies don't like change. Most of the old chunky IBM servers were EOL'd shortly after I started this job.
For the Hardware Calendar error Cisco devices have a Hardware clock and a software clock. They are separate and can be manually configured separately. Try this. You set the clock in enable mode not global config Router#calendar set 07:57:23 23 march 2024 (current time, day, month, year) This will set the hardware calendar clock to whatever you set it too Router#show calendar This output will show the current time, do this to verify it worked Then do the following Router#clock read-calendar This will sync the clock to the calendars time Router#show clock and Router#show calendar should be the same and that error should go away.
@37:30 a reminder of my AS/400, RPG and CL programming days.. fun fact, being of the younger generation in 1996, Windows kids, I wasn't too keen on mono color screens, but we got to appreciate how earlier research had shown that green on black was actually less tiring on the eye than multi coloured displays/UI. Something you come to realise now in your 40s.
5 7/8ozs! 🤣 I completly forgot about the notch differences! I have a vague memory of coming across it, as you did, but it wasn't in a machine I owned. Great to see some more IBM and Cisco goodness!!!! Can't wait for the fiber action and so happy we get to see the interview with your pops! Keep it coming!!
Bro ... Re: your IBM Storage ...I see someone else offered you a DS3000, but some thoughts: *Wild guess* --but that looks like a EXP710 expansion tray; it's not a self-contained complete storage array ... you're going to need (minimum) the following to make it work: - Controller -- that generation is either DS4000 or DS6000 (?); you'll need to cross check (IBM Redbooks?) to see if DS3000 can interop with exp710 expansion trays - Storage Fabric FC (fiber channel) switch (either brocade or Cisco MDS) - fc host adapters (HBA) - fc cables (some spares won't hurt) You might feel the urge to get multiport HBA and that's fine. If you do, you'll also need the corresponding multipath drivers for the respective host OS. IBM used to require an active maintenance/support contract to get access to the drivers and respective microcode ... and with IBM gear, it's like 80% reading docs and prep and 20% keyboard to make it all (eventually) "work".
If, when you start out (assuming you're following the IBM documented guidance) and the performance sucks really bad ... (1) enterprise storage is a mental shift ... its all about scaling for a very specific performance requirement/target (i.e. latency or IOPS or reliability etc ...) and this is often achieved with tricks like multi path connectivity to many "shelves" of high performance FC disks in a specific RAID configuration. (2) have fun performance tuning! The knobs available AND exposed to you along with the available instrumentation are UNPARALLELED! (IMHO Second only to Solaris dtrace) (3) *BE Patient with yourself* Things will get worse before they get better. This is where you need to develop a strong mental framework (I.e. get baselines, document changes, one change at a time, root cause analysis, thorough troubleshooting and clearly identifying where in the stack you are working on at the time) IBM storage gear (and almost equivalently, Sun gear) both taught me so much about systems -- I'm so lucky to have had the opportunity to work with a few of these systems in a large scale enterprise! Seems like you love this stuff, so you're going to have a lot of fun! 😅😊😂
the LICCODE procedure message got me thinking, i wonder if your machine is set for Capacity Use on Demand ("CUoD") ? That's the pay-per-use thing that IBM had that it would lease the hardware (or sell for very very cheap), and the HMC would keep track of your utilisation and call home with that info, and you'd be billed for it. I never really serviced that, only really touched it when we once got a machine setup like that due to some mess up in our order, but the symptom would be pretty much the same, it refused to power on properly (IIRC, because the HMC has to be setup for it specifically) Can you check the licence entitlements on that machine, and see what's in it? Also, it'd be amazing to get this one running, because it (looks like it) does have VIOS support! So you should be able to do full virtualisation on it, down to virtual block devices! (think vmfs-equivalent)
I love these videos. Especially the Cisco stuff. I had my CCNA from 2005-2008 and worked at a company that refurbished and sold old Cisco gear. 2500 routers, 2600xm, pix firewalls, 6500 catalyst switches etc. brings back memories. I’m studying for my CCNA again and I’m buying Cisco gear up lol
SD-RAM, DDR1 and newer, are actually in two groups: unbuffered and buffered. Unbuffered RAM (UDIMM) has no buffer chips. It is compatible with consumer and entry server market, could be non ECC and ECC. Buffered memory usually needs high end server platform, all buffered memory types are ECC. Buffered memory, could be registered RDIMM, from Ddr2 and up also fully buffered FBDIMM or load reduced LRDIMM. They have additional register / buffer chips. There was some exceptions with compatibility, Registered sdram was compatible with desktop 440bx chipsets, registered DDR1 with 939 socket etc.
Also the chassis fans are not field serviceable it's a real pain in the NOC when they die as it's a whole unrack and re rack different chassis. Also the PAEs in the front are not hot swappable from experience non zero chance of shorting and the NPE rebooting or just completely bricking. Neither things you want in a production network lol.
240 V is questionable for 208 I’ve put a few comments an electrician might be able to help you better or some step up transformer for European appliances I think it’s still your suspect on the power supply
Just watching your videos is enough to fulfill my vintage server/networking interests of the day, but also spark some need of buying stuff to play as well. I certainly love your videos man!
Speaking of fiber [sic] channel, we just retired the last FC system at work. It was part of a huge SAN that was also connected to a tape backup library, and was replaced by a much smaller (physically) ISCSI SAN. The new one is all flash, insanely fast, and has way more capacity than the old FC SAN. They were also replacing the entire workflow that ran on the old SAN, which meant restoring hundreds of old backup tapes and transferring the archives into the new system, which archives everything to S3. They've only got a 1Gbps direct connect to the storage gateway in their AWS VPC, but it's still way quicker to restore data from S3 than tape because they don't have to request the tape from the storage company and wait for it to be delivered the next day.
That old school JDE interface screenshot brought back a core memory. A family member managed iSeries gear when I was younger and the company was kind enough to let me hang out in a corner and tinker with the derelict tech from time to time.
My first job after I finished University was at an ISP help desk. Right around the time of ADSL but still a lot of dialup around too. (2003-2004) I got to receive a 7206VXR from our HQ in another state and an ATM card to install j to it before handing off to another team to rack and commission. Good memories 😊
Measuring ram in weight, that is a great one! Can’t wait to see the Cisco stacks grow further. Hopefully you can get the hardware issues resolved and get that P5 system going again.
I have got here something like 5 or 6 boxes of the IBM FC SANs. Tore all but one apart already, gonna use the chassis for *something else, not sure what yet*. The rest has been scrapped fore recycling because who needs FC/SATA limited to 2TB nowadays. Might keep one "just in case of retro" but 3U is just too much to keep around with several boxes of random server hardware.
This is no "special" connector. It is the standard connector for so called FC Drives that where in use from IBM, HP and EMC. EMC used it until the CLARiiON CX4 Series. Then they transitioned to SAS. They even had SAS-FC interposer, so they can use SAS HDD's as replacement for the aging FC HDD's, because the CX4 life was quite extended. The last ones I saw in the 2010's. I installed and designed consumer solutions with those mentioned CLARiiONS for 15 years.
I was surprised when I got to that part of the video, I was expecting the SCA-2 FC-AL connector, which is like a half-size parallel SCSI SCA-80 connector. But nope. Haven't seen that one before.
@@Jerrec OK, I thought I was going insane. You're right, the drives for the FC array in the beginning are sca FC. At the end around 39:22 he pulls a drive out of one of the p5 systems and it's got a card edge hot swap connector I'm unfamiliar with but I guess it's unrelated to FC.
you know your companys hardware is out of date when clabretro starts playing with it. I hate that we're still running 100s of these because no one wants to pay for the replacing them and the software. I've got millions of spare parts for them in my office and I cant get rid of them
Wohoo! Another IBM video dude. And funnily enough, now I'm a datacenter intern (at Bosch) now. Happy to see this content during the weekends when I am chilling and relaxing!
Yeah, Cisco did like hiding those little batteries inside their older gear. I got caught out by an IUPS failure (Interruptible Uninterruptible Power Supply - all UPSes are interruptible!) with a Catalyst 6500, which also took out one of the power supplies. It didn't come back after the support provider replaced the power supply, turned out one of the modules also had a flat battery. Thankfully it was replaceable, because the replacement module that the tech brought was actually faulty, so swapping out the battery brought our old module back. It did require the usual magical CLI incantation to get it booting properly, which the engineer back at the support provider's office knew. They definitely earned their support fees that month, because there's no way I would've been able to get us back up and running that quickly since I've never seen that kind of failure before.
About the HMC RAM - That machine has an unused flot for a second CPU and 8 slots for RAM. So half of those slots are connected to the second CPU, and won't work without it.
As someone whos well versed in the ddr2 (I bought the wrong ram soo many times for my IBm x3650 M1 that I gave up - mainly due to bad eBay listings), I still have the same problem when buying ddr3 for my dell r210 ii and a Fujitsu tower server. I know the pain too well.
One thought: Try removing CPU 0 and then try installing CPU 1 while looking at the connector via the space from CPU 0. That may give you an idea why the second slot won't fully seat. It's possible that the planar board is flexing too much and won't allow the connector pins to slide all the way i, or one end of that monster connector is catching on something. If you stick your phone into the hole while recording video, it may give you a hint.
I should've mentioned in the video but after all the reseating I did while troubleshooting both processor assemblies were seating quite well by the end of it. So unfortunately something wrong with that processor assembly.
If you‘re going down the rabbit hole of FC connectivity, mind compatibility issues. I remember FC in those times you‘re sourcing your hardware from was extremely finicky as far as compatibility was concerned. Sometimes you had to flash one specific version on your HBAs to ensure connectivity between HBAs, GBICs, Switch Ports and JBODs. Ah, the fun we used to have… 🤣
I worked at a datacenter as a technician in high school in the early 2010's, and used to love exploring all the various machines people would bring in to colocate, as well as "junked" / "e-waste" machines we had in the back. We were primarily a supermicro shop but I always thought the PowerEdge servers were the cleanest looking, especially the giant blade chassis. That being said, the supermicros were built like tanks, despite being so barebones/no frills. We had some Supermicro Pentium-4 machines _still_ running at that time (customers who refused to upgrade) and they must have had near 100k of runtime hours at that point. Would love to see an early Blade chassis on the channel at some point, though you'll definitely need a hand lifting it they can easily weigh 200+ lbs loaded.
With the system powered off, your could take that processor assembly PCB off its back plate and check it seats properly on its own. Then you know if it’s a connector issue or a mechanical fit issue.
I wonder if Connor Krukosky would be able to help any. Granted I don't know if this would have anything in common with the IBM z890 he is known for rebuilding. Never hurts to hit him up though. I think he is still active on twitter.
Love your videos and especially the cisco and server stuff. I can't afford the money for the power bill to drive that big machines. But dang they look impressive. Still have a small homeland though :)
Ick, IBM definitely borrowed websphere's UI code. Troubleshooting that and Oracle Weblogic still gives me nightmares, the thing with held together with bubble gum and shoe strings.
This is why I have a strict no rack mounted computers policy at home. I'd just buy a lot of power and sparc servers. Saw this power8 server a couple of years ago that were just $600-$800 shipping included.
This won't protect you, I have 4 Cisco 2811s and a 3560 sitting on a chest of drawers. Ironically I can see a server rack being my next BUI (Buying Under the Influence).
thanks! and definitely, I've got a catalyst 3550 that I'd like to use as the switch for the ibm stuff instead of those little temporary linksys ones. and I'll get some VLANs going on that thing
How interested would you be in an IBM DS3000? It's heavy... can't really afford to ship it to you though :\ But it does have dual PSU, dual controllers, and I think we've got actual caddies for all the disks, plus the rack rails. Originally specced for a 2-node VMware ESXi cluster, each node can connect to both controllers.. I'll even throw in the DAS cables even though that's probably the only thing I could sell 😂 Honestly you could have it for shipping, it'd otherwise just get e-wasted by my company.
hey there... I'd be interested. you can reach out to the email in the channel's about page (might have to be on web to see it). and I'm of course happy to cover shipping!
@@clabretro 😬I'll be in touch, my email is similar to my display name here. I'll get you some pics and specs and the shipping origin and you can decide if you even still want it. I won't be too put out if you end up passing, but I'd love to see it put to good use.. it was built to last, we ran it for **twelve** years.
I don't think that the battery having 3V on it is an indication that the router was being used until very recently, I have a 7206VXR that I got between 2-5 years after decomm and it was running fine, I played with it for a while but it has been sitting on my shelf unpowered for almost another 4 years now, I just checked it and the battery is still at 3.1V, so those things just last a looooong time imo That being said, I always appreciate another vintage cisco shenanigans, I also have a 7507 which is a wonderful piece of hardware (albeit a bit loud and power hungry >.
since you now have 240v, you might as well convert the rest of the rack to use it. as power supplies tend to be slightly more efficient when using 240v directly (in other words, cost savings!).
My only concern and I said it earlier is 208 V AC is Europe and data centers 240 V AC is too hot lines and ground or later too hot lines to ground in a neutral in the US for appliances and unfortunately just wiring it physically is not enough you need a step up transformer something for like European accessories that you could plug your 240 V plug into so I think your power supply is still freaking out because that cable is not able to actually do anything for you Good luck love the videos
Once again, the IBM gods thwart you from demoing LPARS first missing HDD connector. then not enough resources, then not enough power, now a unhappy server I've never seen that "extended DDR2" ever before, I blame IBM lol. I've seen ECC and DDR2 FB DIMMS yes, but not that stuff some thoughts: 1) on the HMC maybe a BIOS flash will expand its QVL/happy list of supported RAM, I've seen BIOS flashes allow for even more RAM than originally intended (happened on the R720 I used to own) yes, I know it changed its mind w the error you where getting the 2nd time but its food for thought. 2) I take it you have tried swapping the CPU modules around in the server and seeing if you still have the "it won't fully seat" issue 3) do the CPU themselves come out? (under the copper heatsink) check for pins (if it has any) in there and maybe reseat-it could be like a modern x86 platform where memory controller is on the CPU itself- if the CPU pops out try switching them around maybe its one of the POWER5+ itself, rather than the entire board 4) try the RAM from the smaller P5 in this one. 5) double-triple check you have the RAM installation order right 6) it might be the VRMs on the server...are they compatible with the smaller working one? yes, I work on Oracle Peoplesoft at work. ITS HORRIBLE AHHHHHHHHH -you can rename a file in MSDOS and it works but in Peoplesoft can you rename-NOPE!
All great thoughts. I did indeed swap the modules and I should've mentioned -- they both seem to seat just fine now. I think the CPUs do come off the assemblies, I'll have to try that. The fins on the broken one do seem a bit banged up so that's suspicious. I'll have to try all the other ideas!
Hiya Clab! Just wanted to say that I really enjoy your videos on these rather, shall we say, vintage, systems! I especially enjoy your Sun Microsystems videos as I have some of that equipment myself and really enjoy it. I have to wonder when you're going to pick up the V880 server though, I got one a couple of years ago in a lot sale for super cheap considering all I got with it. Would make a great edition to your collection though.
Thanks! Yeah definitely always on the lookout for more Sun gear, I really need to get more videos out on that. Would love to get my hands on a V880 someday!
Great vid Clab, glad you got that server at least powered up. Shame about that CPU card, you think it's fixable? I bet your dad's not the only person from Boeing with horror stories.
I got click baited into watching more Cisco VXR content and I'm not even mad about it.
Good for you. I’m looking forward to be clickbaited into watching more Cisco tower of power.
as god said: live electrical work is good ideast for getting in to headven
My company still uses JDE software, running through old AS/400 console emulators... Yeah, big companies don't like change. Most of the old chunky IBM servers were EOL'd shortly after I started this job.
i run the ibm model 3332123 in my ('company')
Just wantedf to go to bed, but you can not not watch a great clabretro video... XD
For the Hardware Calendar error
Cisco devices have a Hardware clock and a software clock. They are separate and can be manually configured separately.
Try this. You set the clock in enable mode not global config
Router#calendar set 07:57:23 23 march 2024 (current time, day, month, year)
This will set the hardware calendar clock to whatever you set it too
Router#show calendar
This output will show the current time, do this to verify it worked
Then do the following
Router#clock read-calendar
This will sync the clock to the calendars time
Router#show clock and Router#show calendar should be the same and that error should go away.
That fixed it! I had tried setting the calendar but didn't actually sync the clock to the calendar time. Thanks!
@37:30 a reminder of my AS/400, RPG and CL programming days.. fun fact, being of the younger generation in 1996, Windows kids, I wasn't too keen on mono color screens, but we got to appreciate how earlier research had shown that green on black was actually less tiring on the eye than multi coloured displays/UI. Something you come to realise now in your 40s.
Aaah, glorious IBM equipment! Case of beers, weekend and tons of cool equipment to play with. 🤘🤩
oh yeah
Next time on Calretro; "My wife fell out of her chair when the power bill came."
with great power come great electricity bill
My thoughts exactly. In Germany you'd go broke running this for just two weeks lol
@@surrealmemes3518with great power com big PEENIS
😂
Not in Ontario since you have to be running it for a month and a half
5 7/8ozs! 🤣 I completly forgot about the notch differences! I have a vague memory of coming across it, as you did, but it wasn't in a machine I owned. Great to see some more IBM and Cisco goodness!!!! Can't wait for the fiber action and so happy we get to see the interview with your pops! Keep it coming!!
haha I thought you'd like the formal weigh in
Bro ... Re: your IBM Storage ...I see someone else offered you a DS3000, but some thoughts:
*Wild guess* --but that looks like a EXP710 expansion tray; it's not a self-contained complete storage array ... you're going to need (minimum) the following to make it work:
- Controller -- that generation is either DS4000 or DS6000 (?); you'll need to cross check (IBM Redbooks?) to see if DS3000 can interop with exp710 expansion trays
- Storage Fabric FC (fiber channel) switch (either brocade or Cisco MDS)
- fc host adapters (HBA)
- fc cables (some spares won't hurt)
You might feel the urge to get multiport HBA and that's fine. If you do, you'll also need the corresponding multipath drivers for the respective host OS. IBM used to require an active maintenance/support contract to get access to the drivers and respective microcode ... and with IBM gear, it's like 80% reading docs and prep and 20% keyboard to make it all (eventually) "work".
If, when you start out (assuming you're following the IBM documented guidance) and the performance sucks really bad ...
(1) enterprise storage is a mental shift ... its all about scaling for a very specific performance requirement/target (i.e. latency or IOPS or reliability etc ...) and this is often achieved with tricks like multi path connectivity to many "shelves" of high performance FC disks in a specific RAID configuration.
(2) have fun performance tuning!
The knobs available AND exposed to you along with the available instrumentation are UNPARALLELED!
(IMHO Second only to Solaris dtrace)
(3) *BE Patient with yourself*
Things will get worse before they get better. This is where you need to develop a strong mental framework (I.e. get baselines, document changes, one change at a time, root cause analysis, thorough troubleshooting and clearly identifying where in the stack you are working on at the time)
IBM storage gear (and almost equivalently, Sun gear) both taught me so much about systems -- I'm so lucky to have had the opportunity to work with a few of these systems in a large scale enterprise!
Seems like you love this stuff, so you're going to have a lot of fun! 😅😊😂
the LICCODE procedure message got me thinking, i wonder if your machine is set for Capacity Use on Demand ("CUoD") ? That's the pay-per-use thing that IBM had that it would lease the hardware (or sell for very very cheap), and the HMC would keep track of your utilisation and call home with that info, and you'd be billed for it.
I never really serviced that, only really touched it when we once got a machine setup like that due to some mess up in our order, but the symptom would be pretty much the same, it refused to power on properly (IIRC, because the HMC has to be setup for it specifically)
Can you check the licence entitlements on that machine, and see what's in it?
Also, it'd be amazing to get this one running, because it (looks like it) does have VIOS support! So you should be able to do full virtualisation on it, down to virtual block devices! (think vmfs-equivalent)
I love these videos. Especially the Cisco stuff. I had my CCNA from 2005-2008 and worked at a company that refurbished and sold old Cisco gear. 2500 routers, 2600xm, pix firewalls, 6500 catalyst switches etc. brings back memories. I’m studying for my CCNA again and I’m buying Cisco gear up lol
SD-RAM, DDR1 and newer, are actually in two groups: unbuffered and buffered. Unbuffered RAM (UDIMM) has no buffer chips. It is compatible with consumer and entry server market, could be non ECC and ECC. Buffered memory usually needs high end server platform, all buffered memory types are ECC. Buffered memory, could be registered RDIMM, from Ddr2 and up also fully buffered FBDIMM or load reduced LRDIMM. They have additional register / buffer chips. There was some exceptions with compatibility, Registered sdram was compatible with desktop 440bx chipsets, registered DDR1 with 939 socket etc.
Mm, buffer chips… 🤤
"Tricked you into watching another VXR Video" we've been had!
Not actually. I quite enjoy these videos, the Cisco stuff never gets old.
Seeing that stack of lInksys routers brings me back to my DD-WRT/Tomato custom firmware days!
Tomato was so much nicer to configure/use than DD-WRT, but it never really took off like DD-WRT did.
I am watching you talk about 720xs while actively decommissioning 720xs in the network I work in.
Also the chassis fans are not field serviceable it's a real pain in the NOC when they die as it's a whole unrack and re rack different chassis. Also the PAEs in the front are not hot swappable from experience non zero chance of shorting and the NPE rebooting or just completely bricking. Neither things you want in a production network lol.
Always good to see another home CTO budget fueled by alcohol. Oddly enough most of my purchasing is on Saturday nights...
funny how that happens
The older networking gear is always fun to learn about.
Nice job getting the 240v hooked up to finally get that p5 up and powered on at least!
Thanks! Forward progress haha
240 V is questionable for 208 I’ve put a few comments an electrician might be able to help you better or some step up transformer for European appliances I think it’s still your suspect on the power supply
Just watching your videos is enough to fulfill my vintage server/networking interests of the day, but also spark some need of buying stuff to play as well. I certainly love your videos man!
that's exactly what I love to hear. thanks for watching!
those were beautiful and reliable machines, runs for years non stop
Speaking of fiber [sic] channel, we just retired the last FC system at work. It was part of a huge SAN that was also connected to a tape backup library, and was replaced by a much smaller (physically) ISCSI SAN. The new one is all flash, insanely fast, and has way more capacity than the old FC SAN. They were also replacing the entire workflow that ran on the old SAN, which meant restoring hundreds of old backup tapes and transferring the archives into the new system, which archives everything to S3. They've only got a 1Gbps direct connect to the storage gateway in their AWS VPC, but it's still way quicker to restore data from S3 than tape because they don't have to request the tape from the storage company and wait for it to be delivered the next day.
That old school JDE interface screenshot brought back a core memory. A family member managed iSeries gear when I was younger and the company was kind enough to let me hang out in a corner and tinker with the derelict tech from time to time.
"Shut up Sun Ray, this is a Cisco channel now!" 😜
poor little Sun Ray
My first job after I finished University was at an ISP help desk. Right around the time of ADSL but still a lot of dialup around too. (2003-2004)
I got to receive a 7206VXR from our HQ in another state and an ATM card to install j to it before handing off to another team to rack and commission.
Good memories 😊
I've got a couple ATM port adapters, I'll eventually try that out too!
Measuring ram in weight, that is a great one! Can’t wait to see the Cisco stacks grow further. Hopefully you can get the hardware issues resolved and get that P5 system going again.
Very nice. I had that happen a few times. You need at least 1 period correct Cisco switch now.
I've got a couple Catalyst 3550s! I'll use one of those instead of those little linksys units
I have got here something like 5 or 6 boxes of the IBM FC SANs. Tore all but one apart already, gonna use the chassis for *something else, not sure what yet*. The rest has been scrapped fore recycling because who needs FC/SATA limited to 2TB nowadays. Might keep one "just in case of retro" but 3U is just too much to keep around with several boxes of random server hardware.
“nothing involved”, watching this amazing channel while havin my boubon
fantastic, just cracked a beer. cheers!
This is no "special" connector. It is the standard connector for so called FC Drives that where in use from IBM, HP and EMC. EMC used it until the CLARiiON CX4 Series. Then they transitioned to SAS. They even had SAS-FC interposer, so they can use SAS HDD's as replacement for the aging FC HDD's, because the CX4 life was quite extended. The last ones I saw in the 2010's.
I installed and designed consumer solutions with those mentioned CLARiiONS for 15 years.
I was surprised when I got to that part of the video, I was expecting the SCA-2 FC-AL connector, which is like a half-size parallel SCSI SCA-80 connector. But nope. Haven't seen that one before.
@@abb1793 I thought that those are exactly those connectors. Hard to measure for me because I dont have those disks at home.
@@Jerrec OK, I thought I was going insane. You're right, the drives for the FC array in the beginning are sca FC. At the end around 39:22 he pulls a drive out of one of the p5 systems and it's got a card edge hot swap connector I'm unfamiliar with but I guess it's unrelated to FC.
@@abb1793 Ok. I was relating to the yellow SCA connector.
you know your companys hardware is out of date when clabretro starts playing with it.
I hate that we're still running 100s of these because no one wants to pay for the replacing them and the software.
I've got millions of spare parts for them in my office and I cant get rid of them
haha. i think it's hard to get rid of IBM once you deploy.
@@clabretro very hard unfortunately
It's Friday night, I've got pizza, whiskey and a new clabretro video just dropped.. BEST FRIDAY EVER!
love it. cheers!
Wohoo! Another IBM video dude. And funnily enough, now I'm a datacenter intern (at Bosch) now. Happy to see this content during the weekends when I am chilling and relaxing!
I think we might still have old UNIX servers in our datacenter, I know we have some AIX servers haha...
More Cisco!
It's so funny to see a small linksys switch powering all of the ethernet connections. Made me laugh lol.
Yeah I'll bet the previous p5 owners wouldn't have thought they'd get hooked up to those haha
@@clabretroMan, I enjoy the RAM rant too man. FBDIMMs vs Registered. It gets me too man, it gets me too.
Those error codes reminds me of Apollo DSKY program alarms. They are constantly changing and maybe five people on entire earth ever knew them all.
Yeah, Cisco did like hiding those little batteries inside their older gear. I got caught out by an IUPS failure (Interruptible Uninterruptible Power Supply - all UPSes are interruptible!) with a Catalyst 6500, which also took out one of the power supplies. It didn't come back after the support provider replaced the power supply, turned out one of the modules also had a flat battery. Thankfully it was replaceable, because the replacement module that the tech brought was actually faulty, so swapping out the battery brought our old module back. It did require the usual magical CLI incantation to get it booting properly, which the engineer back at the support provider's office knew. They definitely earned their support fees that month, because there's no way I would've been able to get us back up and running that quickly since I've never seen that kind of failure before.
Have you ever been asked to rescue an ancient system based on your videos? 😊
not yet!
Great vid as always
thanks!
you are the man Clab! cant wait till you get it fully goin
Finally, a new video! It seemed like forever since the last one!...
i was upset but snow was a viable excuse...this time. Time for a redundant data center!
About the HMC RAM - That machine has an unused flot for a second CPU and 8 slots for RAM. So half of those slots are connected to the second CPU, and won't work without it.
oh geez you're totally right! I didn't even think about that!
Nope, that machine predates the use of integrated memory controllers on Intel systems, so all the RAM slots attach to the northbridge.
@@megabytephreak Probably has two north-bridges.
For a Cisco video.. there's some nice IBM's in there too.
Its nice to see someone else with a stack of those blue and black boxes.
As someone whos well versed in the ddr2 (I bought the wrong ram soo many times for my IBm x3650 M1 that I gave up - mainly due to bad eBay listings), I still have the same problem when buying ddr3 for my dell r210 ii and a Fujitsu tower server. I know the pain too well.
it's the worst
One thought: Try removing CPU 0 and then try installing CPU 1 while looking at the connector via the space from CPU 0. That may give you an idea why the second slot won't fully seat. It's possible that the planar board is flexing too much and won't allow the connector pins to slide all the way i, or one end of that monster connector is catching on something. If you stick your phone into the hole while recording video, it may give you a hint.
I should've mentioned in the video but after all the reseating I did while troubleshooting both processor assemblies were seating quite well by the end of it. So unfortunately something wrong with that processor assembly.
You definitely didn't need to trick me into watching more Cisco content, I appreciate the "2-in-1" video! Best of luck getting that p5 up.
haha thanks
If you‘re going down the rabbit hole of FC connectivity, mind compatibility issues. I remember FC in those times you‘re sourcing your hardware from was extremely finicky as far as compatibility was concerned. Sometimes you had to flash one specific version on your HBAs to ensure connectivity between HBAs, GBICs, Switch Ports and JBODs. Ah, the fun we used to have… 🤣
Hey mah! The IBM guy uploaded a new video :p
Power tower of Cisco gear is growing
I worked at a datacenter as a technician in high school in the early 2010's, and used to love exploring all the various machines people would bring in to colocate, as well as "junked" / "e-waste" machines we had in the back. We were primarily a supermicro shop but I always thought the PowerEdge servers were the cleanest looking, especially the giant blade chassis. That being said, the supermicros were built like tanks, despite being so barebones/no frills. We had some Supermicro Pentium-4 machines _still_ running at that time (customers who refused to upgrade) and they must have had near 100k of runtime hours at that point.
Would love to see an early Blade chassis on the channel at some point, though you'll definitely need a hand lifting it they can easily weigh 200+ lbs loaded.
Sun would never give you this much hassle 😅
I agree!
So many options, they even give you the option to enable RCE's!
If I remember correctly when I was in my high school's it Department we had the HMC Daisy chained
I got an old 19° Rackmount Server just to use it's case for my modern server. It was a work-out getting it up to the 2th floor and out of my car.
clabretro upload. Yes.
With the system powered off, your could take that processor assembly PCB off its back plate and check it seats properly on its own. Then you know if it’s a connector issue or a mechanical fit issue.
I wonder if Connor Krukosky would be able to help any. Granted I don't know if this would have anything in common with the IBM z890 he is known for rebuilding. Never hurts to hit him up though. I think he is still active on twitter.
We still use JDE as our ERP where I work, it's extremely powerful, but boy do I hate it
You are like Peter Seller in the film „the party“ but great you are keep going… 😅
New clabretro let's go!!!!!!!!
No, the DSLAMs are probably the coolest things ever!!!!
Love your videos and especially the cisco and server stuff. I can't afford the money for the power bill to drive that big machines. But dang they look impressive. Still have a small homeland though :)
thanks! I definitely don't leave these old big ones running all the time
Ick, IBM definitely borrowed websphere's UI code. Troubleshooting that and Oracle Weblogic still gives me nightmares, the thing with held together with bubble gum and shoe strings.
That little thinclient there killed me 😂
This is why I have a strict no rack mounted computers policy at home. I'd just buy a lot of power and sparc servers. Saw this power8 server a couple of years ago that were just $600-$800 shipping included.
its dangerous
This won't protect you, I have 4 Cisco 2811s and a 3560 sitting on a chest of drawers.
Ironically I can see a server rack being my next BUI (Buying Under the Influence).
Nice video! Always love seeing old enterprise hardware getting some love. Any plans to mess around with VLANs at some point?
thanks! and definitely, I've got a catalyst 3550 that I'd like to use as the switch for the ibm stuff instead of those little temporary linksys ones. and I'll get some VLANs going on that thing
How interested would you be in an IBM DS3000? It's heavy... can't really afford to ship it to you though :\ But it does have dual PSU, dual controllers, and I think we've got actual caddies for all the disks, plus the rack rails. Originally specced for a 2-node VMware ESXi cluster, each node can connect to both controllers.. I'll even throw in the DAS cables even though that's probably the only thing I could sell 😂 Honestly you could have it for shipping, it'd otherwise just get e-wasted by my company.
hey there... I'd be interested. you can reach out to the email in the channel's about page (might have to be on web to see it). and I'm of course happy to cover shipping!
@@clabretro 😬I'll be in touch, my email is similar to my display name here. I'll get you some pics and specs and the shipping origin and you can decide if you even still want it. I won't be too put out if you end up passing, but I'd love to see it put to good use.. it was built to last, we ran it for **twelve** years.
@@clabretro "I'll have my people get in touch with your people"... 😀
I’m shocked YT suggested this PT 2 to me
I don't think that the battery having 3V on it is an indication that the router was being used until very recently, I have a 7206VXR that I got between 2-5 years after decomm and it was running fine, I played with it for a while but it has been sitting on my shelf unpowered for almost another 4 years now, I just checked it and the battery is still at 3.1V, so those things just last a looooong time imo
That being said, I always appreciate another vintage cisco shenanigans, I also have a 7507 which is a wonderful piece of hardware (albeit a bit loud and power hungry >.
that's a fair point, could've been sitting quite some time. always funny to think support for them ended 7 years ago though haha
You'll get over working on a live panel. Wire the breaker up, then install it. :)
Agreed, seeing how easy it was I'll probably do that next time ha.
awesome, welcome to 240v land 🎉
honestly i was hoping to see you do that 240v thing
yeaaaah I figured there were enough videos covering it already, by people who actually know what they're doing haha.
Well done!
since you now have 240v, you might as well convert the rest of the rack to use it. as power supplies tend to be slightly more efficient when using 240v directly (in other words, cost savings!).
Awesome video. Have you considered using 90s/2000s catalyst switches instead of the linksys for your VLANS?
yup! I'll probably swap those temporary switches out for a Catalyst 3550 with VLANs
My only concern and I said it earlier is 208 V AC is Europe and data centers 240 V AC is too hot lines and ground or later too hot lines to ground in a neutral in the US for appliances and unfortunately just wiring it physically is not enough you need a step up transformer something for like European accessories that you could plug your 240 V plug into so I think your power supply is still freaking out because that cable is not able to actually do anything for you Good luck love the videos
Funnily enough… i found one of these that I think is relatively new in the back room of my current work. I also found the rails if you are interested😂
Interesting! You can shoot me a message at the email in the channel's about section if you'd like. Might have to be on a web browser to see it.
Once again, the IBM gods thwart you from demoing LPARS first missing HDD connector. then not enough resources, then not enough power, now a unhappy server
I've never seen that "extended DDR2" ever before, I blame IBM lol. I've seen ECC and DDR2 FB DIMMS yes, but not that stuff
some thoughts:
1) on the HMC maybe a BIOS flash will expand its QVL/happy list of supported RAM, I've seen BIOS flashes allow for even more RAM than originally intended (happened on the R720 I used to own) yes, I know it changed its mind w the error you where getting the 2nd time but its food for thought.
2) I take it you have tried swapping the CPU modules around in the server and seeing if you still have the "it won't fully seat" issue
3) do the CPU themselves come out? (under the copper heatsink) check for pins (if it has any) in there and maybe reseat-it could be like a modern x86 platform where memory controller is on the CPU itself- if the CPU pops out try switching them around maybe its one of the POWER5+ itself, rather than the entire board
4) try the RAM from the smaller P5 in this one.
5) double-triple check you have the RAM installation order right
6) it might be the VRMs on the server...are they compatible with the smaller working one?
yes, I work on Oracle Peoplesoft at work. ITS HORRIBLE AHHHHHHHHH -you can rename a file in MSDOS and it works but in Peoplesoft can you rename-NOPE!
All great thoughts. I did indeed swap the modules and I should've mentioned -- they both seem to seat just fine now.
I think the CPUs do come off the assemblies, I'll have to try that. The fins on the broken one do seem a bit banged up so that's suspicious.
I'll have to try all the other ideas!
Hiya Clab! Just wanted to say that I really enjoy your videos on these rather, shall we say, vintage, systems! I especially enjoy your Sun Microsystems videos as I have some of that equipment myself and really enjoy it. I have to wonder when you're going to pick up the V880 server though, I got one a couple of years ago in a lot sale for super cheap considering all I got with it. Would make a great edition to your collection though.
Thanks! Yeah definitely always on the lookout for more Sun gear, I really need to get more videos out on that. Would love to get my hands on a V880 someday!
Are you planning to get IBM I running on that thing? It's a fun odd system.
unfortunately neither p5 is licensed to run IBM I, but I definitely want to dive into IBM I in the future!
IBM P5 and Cisco...win win! nice video.
thanks!
You have a very interesting channel sir!
thanks!
P.S. have you swapped the processor modules over to see if the problem moves too?
yeah it's always the one assembly regardless of slot
Cisco MPE makes me feel like GOD
There's no kill like overkill..!
next sat ill be taking possession of a cisco 6500 fully populated, sadly only gb but ill play with it then just get rid of it, its heavy lol
Thank you for including real numbers along with the gibberish when weighing your RAM ❤ 😂
You, sir, have a serious 72XX problem. You should look into something like a Juniper M20 or one of the smaller MX series.
I can't stop
Great vid Clab, glad you got that server at least powered up. Shame about that CPU card, you think it's fixable?
I bet your dad's not the only person from Boeing with horror stories.
How much power your server needs?
-Yes
Very cool!
22:55 A protogen would never go hungry :O
Ponies mate, linksys ponies and cisco horses :P
😂
I'm counting 14 instead of 16 bays in the FibreChannel Array, so you may end up buying two spares when you order the entire population...
yeah I dunno why I said that, it's 14 total and I have 7 lol
I nearly got nerd sniped by some sparc T-1 and T-2 servers on ebay after watching your videos.
close call
not sure about these Cisco screws, but they look if somebody already messed with them. They appear like inbus screws.
LINKSYS STACK GANG
New video!! Wooohooo🎉🎉
"...that would be overkill!" 🤣🤣🤣
Yes!
Have you ever "found: and flashed the latest firmware on these old routers?
I have tons of copies of various firmware versions but haven't actually tried upgrading any. Will do that eventually!
oh yeah, more three letter vids