|2012-11-02 22:36:00||pissed off||Amused to death – Roger Waters|
It’s Friday afternoon, I’m in Napier supposedly to do some research work before going to a ‘how to do your MSc dissertation’ lecture. However, I’m in no fit state to work, so I’m going to try to blog some bile out. You have been warned.
Towards the end of last week, Iggy started hanging randomly every 30 minutes or so. There was no detectable set of circumstances to the hangs, so I began to fear the OS was hosed. The system logs mentioned some issues with fonts, so I removed all the fonts in my user library. Still lots of hanging. Eventually, I decided to nuke and pave Iggy’s boot SSD – if there was a software problem, reinstallation from scratch would remove it. If there was a hardware problem, it would presumably continue to manifest itself.
It may help to know that Iggy has 4 storage devices:
- in drive bay 1: 640GB 3·5-inch rotating hard disk, containing nightly CarbonCopyCloner clone of boot SSD (drive bay 4)
- in drive bay 2: 1TB 3·5-inch rotating hard disk, partitioned into two 500GB volumes
- TimeMachine back-up of boot SSD
- CCC clone of MacOS 10.6 (‘SnowLeopard’)
- in drive bay 3: 170GB 2·5-inch rotating hard disk, containing various VirtualBox virtual machines
- in drive bay 4: 512GB 2·5-inch SSD, containing MacOS 10.7.5 (‘Lion’), apps and data.
So with all backups (local backups via TimeMachine to a TimeCapsule and a NAS, CarbonCopyCloner to HD in bay 1, Crashplan offsite) refreshed, I erased, overwrote with zeros and repartitioned Iggy’s SSD, then installed Lion. (The erase and over-write stage to a few hours.) The hangs recurred, so I definitely had a hardware issue. The system logs were very unpleasant reading indeed.
I had some freelance work to do, so I did that while running Iggy from the CarbonCopyClone on the HD in bay 1. During this time, the SSD disappeared from the Finder, prompting a message not to improperly remove drives. The SSD didn’t reappear in the Finder, and was invisible to DiskUtility.
The next step was to find what bit of Iggy was ill. It could have been
- the SSD itself
- the IcyDock that connects it to the the hard drive bay
- drive bay 4 itself
- some other part of Iggy entirely.
To eliminate the last option, I removed the SSD and IcyDock, and ran Iggy from the CarbonCopyClone in bay 1 for an hour, including booting and doing some stuff in all the VMs that live in bay 3. No problems occurred, so all of Iggy, apart from possibly drive bay 4, was OK.
To check drive bay 4, I put the HD containing the VMs into it. I then ran the Windows 7 VM, installed iTunes in it, downloaded a 30-minute podcast and played it twice over. No problems occurred, so drive bay 4 was OK.
To check the IcyDock, I put the HD containing the VMs into it, the repeated the ‘podcast’ test. No problems occurred, so the IcyDock was OK. This meant that the SSD was the culprit. It was under a year old, and had a 3-year warranty. I also had the email confirming the order for the SSD, so I was fairly confident I’d be able to obtain a replacement.
Crucial, the SSD’s manufacturer, said they would be willing to replace the SSD but first I’d need to try reconditioning it. To do this, I needed to connect the SSD to a shut-down mac, restart it as far as the EFI boot-device selector and then leave it in this state overnight. I duly did so, over Monday night.
The next day, I put the SSD back into Iggy. The SSD appeared in the Finder as normal, and I was able to install Mac OS. However, another hang occurred when updating from MacOS 10·8 to 10·8·2, and again when copying back my data. Obviously the SSD wasn’t fixed. I called Crucial again: they advised a process of
- Attaching the SSD to a mac
- booting as far as the EFI boot-device selector
- leaving the mac and SSD in this state for 30 minutes
- shutting down the mac
- disconnecting the SSD and leaving it unpowered for a minute
- repeating steps 1–5 twice
- updating the SSD’s firmware.
The latter step involved downloading an ISO from Crucial, burning it to a CD, booting Iggy from that and letting the firmware software work its magic. After this, the SSD seemed stable, so I began the long haul of reinstalling OS and apps, copying back data, configuring apps (especially setting up Creative Suite 5 to produce print-ready PDFs to the format that my freelance work requires).
A day later, Iggy was back in action, now running MacOS 10·8·2 (‘Mountain Lion’) and running sweet and smart. SSDs are so much faster than rotating HDs! There is 30GB more unused space on the SSD than before – I guess this is due to not reinstalling some apps I just don’t use, and other apps not having gone through so many update stages. Here ended Wednesday.
I woke up during Wednesday night due to a severe hypo – I guess I’d not eaten enough after Wednesday evening’s spinning. So I staggered through Thursday’s lectures and a meeting about publishing my Community Council research. I then did some reading about the style of report I’d need to produce, then headed home via Boots to collect some medication I needed. On the way, Lev Davidovitch’s pannier fell off, thus giving me palpitations about the health of my MacBook Air. (My fears weren’t realised – I’m typing this dreary prose on it.)
At Boots, It turned out that my doctor hadn’t returned the renewed prescription to the store, so they contacted the surgery to get them to fax it over. I also went to the surgery to try to arrange an appointment to discuss some other issues. The surgery’s appointment system is weird: the next available pre-bookable appointment would be 12th November, but if I phoned the following morning (2nd November) at 8am, I’d be able to get an appointment that day.
I then went to the Angus Millar lecture but slept through most of it. After the lecture, I realised I was very hungry, and wasn’t surprised that my my blood sugar level was under 3 mmol l–1. A quick curry at the Mosque Kitchen fixed that. However, this wasn’t the end of blood-sugar issues.
Because I’d slept earlier, I didn’t get to sleep until about 2am, only to wake an hour later due to another hypo. I also woke at 6 with yet another hypo. So I then missed waking up in time to call the surgery for a same-day appointment. I had a fairly unpleasant headache, so I swallowed some ibuprofens, made breakfast and bimbled about until my head was clear enough to venture into the day.
My first port of call was the surgery, to drop off a full sharps bin. I was waiting for a gap in the traffic so I could cycle away from the surgery when a car reversed along the kerb and pushed into me. It had been stationary, about 2 metres away from me and I’d been waiting in the same position for a couple of minutes for a suitable gap in the traffic. So during all that time, the driver hadn’t checked his or her path was clear! I wear a very bright cycling jacket and helmet cover, so I should be visible to anyone who actually looks.
I don’t claim to be perfect in any way but my lapses aren’t potentially fatal, certainly not to anyone else. (The approved order is (1) check behind before signalling, (2) signal while facing forward, (3) get both hands back on handlebars, (4) lifesaver-check behind, (5) manoeuvre; with rules about which side you should check. My signalling technique is adequate but not perfect, mostly due to my shoulder-mobility issues, and I tend to check behind while signalling. [This does not mean I don’t check in front of me before manoeuvring!])
Don’t read the following if you are squeamish about digestive malfunctions.
A few other errands later, I arrived at Napier. By now it was lunchtime. While eating, I suddenly felt nature call me from behind, very urgently. I’ve yet to understand why there is only male toilet near the canteen – I didn’t feel confident about the extra 30 metres to the bigger set of toilets in the computer centre. As it was, I leaked a bit on the way to the nearby toilet. A few unpleasantly long, liquid events later, I made two discoveries:
- the leak had soiled my cycle-shorts
- there was no toilet paper in the cubicle!
I wiped with a pair of underpants that happened to be lurking in the bottom of my pannier. By the end of this, they were in a state I don’t want to recall, let alone try to describe. I didn’t dare try to flush them away, and there was no bin in the toilet area. Fortunately, I had a ziplock bag in my pannier, so I used that to bag up the moist smelly item, and then gingerly waddled away to find a last resting place for my pants.
I would have gone home to change but I was due at a lecture about doing MSc dissertations an hour later. If I’d gone home, I’d not have returned for the lecture. This lecture was informative but somewhat frightening. Afterwards, I was chatting with a classmate when I passed wind – and followed through, in an unpleasantly liquid manner. I can’t be more grateful for the noise- and liquid-absorbance of cycle-shorts.
Some cleaning later, I hesitantly mounted Lev and cycled home. There’s something quite nasty about cycling in soiled shorts. After a shower and a bath, and washing all the clothes I’d been wearing, my hunger was thwarted by rather high blood sugar (17 mmol l–1). I’m now lying in bed with my MacBook Air, feeling my stomach churn, not daring to go to sleep in case I soil the bed.